OSWERHCHAP
HANDBOOK
OF
CHEMICAL HAZARD ANALYSIS
PROCEDURES
FEDERAL EMERGENCY MANAGEMENT AGENCY
f i,
U.S. DEPARTMENT OF TRANSPORTATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
document has ceen reviewed by the Federal Emergency Management
tjiegepartment of* ansportaFion, and the Environmental
~~
£ReQiiteabs_necessiJ[y re£ lect the views ............ and ........ polices
.— "
-_.
- an
of tradenames
^
o"recommenaation for use
-------
FOREWORD
Introduction
The Federal Government has a long record of concern about hazardous materials and their
potential impact on people and the environment Over the years, several Federal agencies
have provided training, technical assistance and guidance to State and local governments and
industry in planning and response to emergencies involving hazardous materials. For
example, the Federal Emergency Management Agency (FEMA) published the Planning
Guide and Checklist for Hazardous Materials Contingency Plans (FEMA-10) in 1981 to
assist communities developing emergency response plans The Department of Transporta-
tion (DOT) has published several editions of the Emergency Response Guidebook to serve as
guidance for initial action to be taken by fire fighters, police, and emergency services
personnel at the scene of transportation incidents involving hazardous materials In 1985, the
Environmental Protection Agency (EPA) published Chemical Emergency Preparedness
Program (CEPP) Interim Guidance to provide technical assistance for a voluntary program
focusing on airborne toxic chemicals under EPA's National Strategy for Toxic Air
Pollutants
Government-wide guidance on emergency planning for hazardous material was introduced in
1987 after the passage of Title in of the Superfund Amendments and Reauthonzation Act
(SARA) with the publication of the National Response Team's Hazardous Materials
Emergency Planning Guide (NRT-1). This effort to coordinate Federal planning processes
concerning specific hazardous materials addressed by SARA was followed with the joint
publication by EPA, FEMA and DOT of Technical Guidance for Hazards Analysis
Handbook of Chemical Hazards Analysis Procedures
This Handbook of Chemical Hazard Analysis Procedures has several objectives one of which
is to expand NRT-1 and the Technical Guidance on Hazards Analysis document by including
information for explosive, flammable, reactive and otherwise dangerous chemicals Al-
though NRT-1 was aimed at addressing planning for all types of hazardous materials, SARA
Title m required local planners to focus on a specific initial list of acutely toxic chemicals
(referred to as Extremely Hazardous Substances) due to their high inhalation toxicity when
airborne, and this was the primary focus of the supplemental guidance document By
introducing additional methodologies on how to plan for these and other dangerous
chemicals, this handbook serves as a stepping stone from NRT-1 and the Technical Guidance
on Hazards Analysis to a more comprehensive approach for emergency planning If deemed
necessary and appropriate by the National Response Team after distribution and field use of
this handbook by emergency planning personnel, a further enhanced hazard analysis guide
may be prepared in the future.
Because this handbook provides methods to investigate local hazards in greater detail than
permitted by earlier guidance, results of calculations using air dispersion models may differ
The Federal Government is continuing to evaluate these types of models and others to
determine the degree of impact on calculations concerning the consequences of a chemical
release For these reasons and because this handbook requires use of the accompanying
software for full utilization, users should carefully assess accident scenarios selected for
evaluation to ensure that computational procedures are appropnate for the chemical being
-------
studied. Difficulties encountered and suggestions or comments (both positive and negative)
should be submitted to DOT, FEMA, and/or the EPA. Be advised that workshops are being
planned by these Agencies during 1989 to address comments, gather input on the handbook
and the related software, and explain their functions Similarly, DOT, FEMA and EPA are
interested in receiving information on problems and expenences associated with use of the
Technical Guidance on Hazards Analysis document and NRT-1.
Beyond providing additional methodologies for assessing the potential impacts of hazardous
material releases, this handbook also expands the three-step hazards analysis approach
(hazard identification, vulnerability analysis, and risk analysis) presented in NRT-1 and its
supplement by introducing a four-step approach involving hazard identification, consequence
analysis, probability analysis, and risk analysis In addition, it provides a tutorial on
hazardous chemicals, suggestions for applying hazard analysis results to writing and updating
an emergency plan, and an expanded discussion of issues relating to sheltenng-in-place
(in-place protection) and evacuation. Because additional projects are underway concern-
ing these and other topics described in Chapter 14 and Appendix C of the handbook,
sponsoring agencies are especially interested in comments on these sections. The
•workshops mentioned above will provide an opportunity for discussion and comment.
General comments on the handbook, its associated computer program named
ARCHIE, and earlier planning aids are highly encouraged and may also be submitted
in writing to:
Federal Emergency Management Agency
Technological Hazards Division
Federal Center Plaza
500 C Street, S.W
Washington, DC 20472
U.S. Department of Transportation
Research and Special Programs Administration
Office of Hazardous Materials Transportation
DHM-50,400 7th Street, S.W.
Washington, DC 20590
U.S Environmental Protection Agency
Chemical Emergency Preparedness and Prevention Office
401M Street, S.W., OS-120
Washington, DC 20460
Alternatively, input to these agencies may be transmitted via use of the Hazardous Materials
Information Exchange (HMIX) computerized bulletin board system operated and maintained
by FEMA and the DOT. HMIX includes a Conference dedicated to ARCHIE where users
may leave messages, questions or comments relating to the program or handbook, exchange
viewpoints, and receive responses to inquiries HMIX may be accessed by modem and
commercial phone line af
(312) 972-3275
-------
An HMDC users manual and technical assistance can be obtained by calling-
1-800-PLAN-FOR Nationwide
or
1-800-367-9592 in Illinois
If you are unable to access HMIX to submit comments or questions relating specifically to
the computer program, please send them in writing to*
ARCHIE Support (DHM-51/Room 8104)
Office of Hazardous Materials Transportation
Research and Special Programs Administration
US Department of Transportation
400 7th Street, SW
Washington, D.C. 20590
Additional copies of this handbook maybe obtained by writing to
Federal Emergency Management Agency
Publications Office
500 C Street, SW.
Washington, D.C. 20472
-------
-------
TABLE OF CONTENTS
1.0 Introduction 1-1
1.1 Background 1-1
1 2 Related Planning Guides and Documents 1-2
2.0 Key Properties of Chemical Substances 2-1
21 States of Matter 2-1
2.2 Definitions of Temperature and Heat 2-2
23 Definition of Pressure 2-3
2 4 Vapor Pressures of Liquids and Solids 2-4
2 5 Boiling Points as a Function of Pressure 2-9
2 6 Definitions of Specific Gravity and Density 2-10
27 Solubility in Water 2-13
2.8 Molecular Weights of Chemical Substances 2-14
3.0 Actions Upon Release to the Environment 3-1
31 Physical State Prior to Release 3-1
3.2 Material States During and Initially After Release 3-1
3 3 Discharges Onto Land 3-5
3 4 Discharges Into Water 3-7
3 5 Fundamental Concepts Pertaining to Discharges into Air 3-11
3.6 Variables that Influence Atmospheric Vapor Dispersion 3-17
4.0 Fire Hazards of Chemical Substances 4-1
4.1 Introduction 4-1
4 2 Measures of Flammability Potential 4-1
43 Measures of Flammability Effects 4-6
44 Types of Fires 4-6
4 5 Products of Combustion 4-11
5.0 Explosion Hazards of Chemical Substances 5-1
51 Definitions 5-1
5 2 Factors that Influence Explosion Potential 5-2
-------
TABLE OF CONTENTS (Cont.)
Page
53 Measures of Explosion Effects 5-3
5.4 Types of Explosions 5-7
6.0 Toxicity Hazards of Chemical Substances 6-1
6.1 Introduction 6-1
6.2 Routes of Entry 6-1
6.3 Types of Toxic Effects 6-3
6.4 Acute Vs Chronic Hazards 6-6
6.5 Importance of Exposure Level and Duration 6-7
6.6 Toxicity Vs. Toxic Hazard 6-8
6.7 Recognized Exposure Limits for Airborne Contaminants 6-9
6.8 Advantages and Disadvantages of Various Limits 6-16
6.9 Relationship of Recommendations to EPA LOCs 6-18
6.10 Consideration of Mixtures of Harmful Gases and Vapors 6-19
6.11 Exposure Limits for Contaminated Water 6-21
6 12 Understanding Toxicological Data in the Literature 6-22
7.0 Reactivity Hazards of Chemical Substances 7-1
7.1 Introduction 7-1
7.2 Exothermic Reactions 7-2
7.3 Neutralization Reactions 7-4
7.4 Corrosivity Hazards 7-5
7.5 Other Hazardous Results or Products of Reactions 7-6
7.6 Sources of Chemical Reactivity Data 7-6
8.0 Hazardous Material Classification Systems 8-1
8.1 Introduction 8-1
8.2 U.S Department of Transportation Classifications 8-1
8 3 U.S. Environmental Protection Agency Classifications 8-5
8.4 National Fire Protection Association Hazard Rankings 8-8
8 5 International Maritime Organization Classifications 8-8
-------
TABLE OF CONTENTS (Cont.)
Page
9.0 Overview of the Hazard Analysis Process 9-1
9.1 Introduction 9-1
92 Step 1: Hazard Identification 9-3
93 Step 2 Probability Analysis 9-4
94 Step 3: Consequence Analysis 9-5
95 Step 4- Risk Analysis 9-5
96 Step 5 Use of Hazard Analysis Results in Emergency Planning 9-6
10.0 Hazard Identification Guidelines 10-1
101 Introduction 10-1
10.2 Reason for the Desired Information 10-1
10 3 Suggested Scope of the Effort 10-2
10.4 Nature of Desired Information 10-4
105 Available Methodologies to Compile Desired Information 10-12
106 Sources of Additional Hazard Identification Guidance 10-20
10.7 Formulation of Credible Accident Scenarios for
Planning Purposes 10-20
10 8 Organization of the Data 10-21
11.0 Probability Analysis Procedures 11-1
11.1 Introduction 11-1
112 General Seventy of Accidents Considered 11-3
113 Bulk Transportation of Hazardous Matenals by Highway 11-3
114 Bulk Transportation of Hazardous Matenals by Rail 11-13
115 Bulk Transportation of Hazardous Matenals by
Marine Vessels 11-20
116 Transportation of Hazardous Matenals by Pipeline 11-26
117 Handling and Transfer of Hazardous Matenals at
Fixed Facilities 11-31
118 Transportation of Packaged Hazardous Materials 11-40
119 Transportation of Hazardous Matenals by Air 11-42
11 10 Summary 11-43
1111 References 11-43
-------
TABLE OF CONTENTS (Cont.)
Page
12.0 Consequence Analysis Procedures 12-1
12.1 Introduction to ARCHIE 12-1
12.2 Installation of the ARCHIE Computer Program 12-6
12.3 General Notes on Responding to Questions from the Program 12-7
12.4 Initialization of Program Configuration Settings 12-9
12 5 Display of the Program Title Screen 12-10
12.6 Introduction to Options on the Main Task Selection Menu 12-10
12.7 Introduction to the Hazard Assessment Model Selection Menu 12-14
12.8 Discharge Menu Option A- Non-Pressurized Rectangular
Tank of Liquid 12-20
12.9 Discharge Menu Option B: Non-Pressurized Spherical Tank
of Liquid 12-23
12.10 Discharge Menu Option C: Non-Pressurized Vertical
Cylinder of Liquid 12-26
12.11 Discharge Menu Option D: Non-Pressurized Horizontal
Cylinder of Liquid 12-29
12.12 Discharge Menu Option E- Pressurized Liquid when Discharge
Location is 4 Inches or Less from the Tank Surface 12-32
12.13 Discharge Menu Option F Pressurized Liquid when Discharge
Location is More Than 4 Inches from the Tank Surface 12-36
12 14 Discharge Menu Option Gf Pressurized Gas Release from
any Container 12-40
12.15 Discharge Menu Option H Release from a Pressurized
Liquid Pipeline 12-44
12 16 Discharge Menu Option I* Release from a Pressurized
Gas Pipeline 12-49
12.17 Hazard Model Menu Option B Pool Area Estimation Methods 12-52
12.18 Hazard Model Menu Option C Pool Evaporation Rate
and Duration Estimates 12-56
12.19 Hazard Model Menu Option D Toxic Vapor Dispersion Model 12-59
1220 Hazard Model Menu Option E. Liquid Pool Fire Model 12-63
12.21 Hazard Model Menu Option F Flame Jet Model 12-64
1222 Hazard Model Menu Option G Fireball Thermal
Radiation Model 12-67
-------
TABLE OF CONTENTS (Cont.)
12.23 Hazard Model Menu Option H Vapor Cloud or Plume
Fire Model 12-68
12.24 Hazard Model Menu Option I: Unconfined Vapor Cloud
Explosion Model 12-72
1225 Hazard Model Menu Option J. Tank Overpressurization
Explosion Model 12-75
1226 Hazard Model Menu Option K. Condensed-Phase
Explosion Model 12-78
12.27 Remaining Options on the Hazard Assessment Model
Selection Menu 12-80
12 28 Use of the Vapor Pressure Input Assistance Subprogram 12-81
12 29 Use of the Tank and Container Contents
Characterization Subprogram 12-84
1230 Other Computer Programs 12-85
13.0 Formulation of a Planning Basis 13-1
131 Introduction 13-1
13 2 Definition of Annual Accident Probability Categories 13-1
13 3 Definition of Accident Seventy Categones 13-3
13 4 Application of Screening Guidelines 13-5
135 Motivation for Continued Planning 13-7
14.0 Use of Hazard Analysis Results In Emergency Planning 14-1
141 Introduction 14-1
14 2 Organization of the Chapter 14-2
14.3 Additional Sources of Planning Guidance and Information 14-2
Appendix A: A Tutorial on Fundamental Mathematical Skills A-1
Appendix B Technical Basis for Consequence Analysis Procedures B-l
Appendix C Overview of "Shelter-m-Place" Concepts C-l
Appendix D Chemical Compatibility Chart D-l
Appendix E Guide to Installation of the ARCHIE Computer Program E-l
Appendix F Basis of Probability Analysis Procedures F-l
V
-------
LIST OF TABLES
Table Page
2.1 Vapor Pressures as a Function of Temperature 2-6
2 2 Atomic Weights of Common Elements 2-15
3.1 Typical States of Materials in Storage or Transportation
Containers 3-2
3 2 Expected Behavior of Spills Into Water 3-8
3.3 Atmospheric Stability Class Selection Table 3-19
4.1 Example Flammability Characteristics 4-5
4.2 Thermal Radiation Burn Injury Catena 4-7
5.1 Explosion Overpressure Damage Estimates 5-5
6.1 Effects of Oxygen Depletion 6-5
6.2 Four Stages of Asphyxiation 6-5
6 3 Summary of Emergency Exposure Guidance Levels from the
National Research Council 6-14
8.1 NFPA Hazard Rankings 8-9
8 2 Basic IMO Material Classes and Divisions 8-11
10.1 Spill Source Characterization Factors 10-5
10 2 Miscellaneous Potential Spill Sources 10-11
11.1 Chemical Accidents Requiring Evacuations 11-4
11.2 Estimated Number of Transportation Accidents 11-4
11.3 Major Hazardous Materials Accidents 11-5
11.4 Suggested Figures for Truck Transportation 11-9
11.5 Suggested Figures for Rail Transportation 11-17
11.6 Suggested Figures for Marine Transportation 11-23
11.7 Suggested Figures for Pipeline Transportation 11-29
11.8 Suggested Figures for Fixed Facilities 11-36
12.1 Special Index to Chapter 12-
12.2 Main Task Selection Menu 12-
12.3 System Description and Use Instructions 12-
12.4 Hazard Assessment Model Selection Menu 12-
12.5 Discharge Model Selection Menu 12-
VI
-------
LIST OF TABLES (Cont.)
Table Page
12 6 Assistance Display for Liquid Specific Heat 12-39
12.7 Assistance Display for Vapor/Gas Specific Heat Ratio 12-39
12 8 Atmospheric Stability Class Selection Table 12-54
12 9 Assistance Display for Explosion Yield Factor 12-74
12 10 Example Output from Explosion Models 12-76
13 1 Annual Individual Mortality Rates for Natural and Accidental
Causes of Death 13-8
14 1 Index to Planning Items 14-6
LIST OF FIGURES
Figure Page
2 1 Evaporation and Vapor-Liquid Equilibrium Phenomena 2-7
3.1 Initial Stages of Vapor Cloud or Puff Dispersion 3-13
3 2 Cross-Section of Puff Concentration Vs. Time 3-14
3.3 Maximum Puff Concentration Vs. Time or Distance 3-14
3 4 Puff or Cloud Isopleths at Increasing Tunes 3-15
3 5 Isopleths in a Continuous Plume 3-15
3 6 Behavior of Lighter Than Air Puffs or Clouds 3-21
3 7 Some Effects of Elevated Emissions 3-23
3 8 Vapor Dispersion Hazard Zone Boundaries 3-27
8.1 US DOT Placards 8-6
12 1 Overview of Acute Spill Hazards on Land 12-16
12 2 Model Selection Charts 12-17
13.1 Accident Frequency/Seventy Screening Matrix 13-6
vii
-------
-------
1.0 INTRODUCTION
1.1 BACKGROUND
The fact that hazardous materials pose a threat to public safety and the environment is
of vital concern to industry and all levels of government, particularly in the aftermath of the
tragedy in Bhophal, India, that took over 2000 lives and injured tens of thousands of others in
the course of a few hours. Although the safety record of the oil and gas and chemical
manufacturing and transportation industries in the United States has been excellent in recent
years, and there has not been a similar catastrophic accident or incident with major loss of
life in the United States in several decades, there is nevertheless a clear need for constant
vigilance on the part of government agencies and those responsible for the movement and
handling of hazardous materials to minimize the possibility of significant discharges to the
external environment. Similarly, there is a clear and possibly even more urgent need to
ensure that both government and industry are prepared to respond quickly, efficiently, and
effectively in the event of an accident to reduce or prevent adverse impacts on public safety
and the environment Tune is critical in the first moments of an accident. A mismanaged
response due to a lack of preplanning can contribute to the causation of fatalities and injuries
as well as an increase in damage to property and the environment
The primary purpose of this handbook and its associated computer program is to
provide emergency planning personnel with the resources necessary to undertake comprehen-
sive evaluations of potentially hazardous facilities and activities within their respective
jurisdictions and thereby formulate a basis for their planning efforts Chapters 2 through 8 of
the handbook discuss fundamental definitions and concepts relating to hazardous material
properties and associated threats to public safety. Chapter 9 provides an overview of the
overall hazard analysis process required to identify, characterize, and evaluate the subject
threats Chapter 10 follows with specific guidance relating to hazard identification while
Chapter 11 provides assistance in evaluating the likelihood that any given accident or
incident will actually occur in the foreseeable future Chapter 12 describes and discusses the
Automated Resource for Chemical Hazard Incident Evaluation (ARCHIE) computer
program and how it may be used to conduct consequence analysis for postulated accident
scenarios Chapter 13 next guides the user through a simplified nsk analysis procedure
designed to provide a planning basis, while Chapter 14 provides guidance on how results of
the overall hazard analysis process may be utilized in development of a comprehensive
emergency response plan
Several appendices to the handbook provide additional guidance and details Appendix
A is a tutorial on fundamental mathematical skills Appendix B presents an overview of the
technical basis for consequence analysis procedures, while Appendix C provides an overview
of "Shelter-m-Place" concepts. Appendix D follows with the presentation of a chemical
-------
compatibility chart for potentially reactive materials. Appendix E is a guide to installation of
the ARCHIE computer program, while Appendix F ends the handbook with the basis for
accident/Incident probability analysis procedures
12 RELATED PLANNING GUIDES AND DOCUMENTS
Multi-Agency Publications of the Federal Government
The National Response Team (NRT) is comprised of representatives of 14 federal
agencies having major responsibilities for issues involving the environment, transportation,
and public health and safety. It is the primary body in the United States charged with
responsibility for planning, preparedness, and response actions related to spills or discharges
of oil and hazardous materials into the environment.
The NRT published the Hazardous Materials Emergency Planning Guide in March
1987 as document NRT-1. This guide provides a fairly detailed overview of the efforts
required for:
• Selecting and organizing an emergency planning team
• Defining the tasks of the planning team
• Developing an emergency plan and individual plan elements
• Appraising, testing, and maintaining the plan
The guide focuses on the needs and requirements of public authorities in local and state
governments but also contains useful information for industrial planning personnel in terms
of the basic elements of the planning process. Additionally, it provides insights into those
issues of concern to public authorities and the importance of cooperation and coordination of
emergency planning activities between the public and private sectors. Copies of the guide
are available by writing:
Hazardous Materials Emergency Planning Guide
OS-120
401 M Street, S.W.
Washington, D.C. 20460
Subsequent to completion and distribution of NRT-1, the U.S. Environmental Protec-
tion Agency (EPA), in conjunction with the Federal Emergency Management Agency
(FEMA) and the U.S. Department of Transportation (DOT), published Technical Guidance
for Hazards Analysis — Emergency Planning for Extremely Hazardous Substances to
fulfill obligations mandated by the Superfund Amendments and Reauthonzation Act (SARA)
of 1986. Focusing primarily on the hazards associated with a specific list of highly toxic
substances deemed to pose acute inhalation hazards when discharged into the atmosphere,
1-2
-------
the guide provides simplified guidance for hazard identification, vulnerability analysis, and
risk analysis of facilities subject to reporting requirements under Title in of SARA.
Additionally, the document contains a simplified screening procedure for ranking the threats
due to designated Extremely Hazardous Substances (EHS) in a community Copies may be
obtained by writing the same address given above for NRT-1.
Publications of the Federal Emergency Management Agency
The Federal Emergency Management Agency (FEMA) publishes the Guide for
Development of State and Local Emergency Operations Plans (CPG 1-8) and the Guide for
Review of State and Local Emergency Operations Plans (CPG 1-8A), which provide
information to emergency management personnel and state and local government officials
about FEMA's concept of planning under the Integrated Emergency Management System
(IEMS). This system emphasizes integration of planning for all types of hazards that pose a
threat to a community and provides extensive guidance in the coordination, development,
review, validation, and revision of emergency operations plans.
The concepts if not the specific details of FEMA's guidance are applicable to
individual communities and chemical facilities in that many such sites may be subject to a
variety of natural and technological hazards. Under a wide variety of circumstances, a single
emergency plan that provides "umbrella coverage" for a locality can ensure increased
efficiency and effectiveness of a planning effort by reducing duplication of common
activities.
FEMA, in conjunction with DOT and the EPA, has also published a wide variety of
emergency planning guidance documents relating to emergencies involving nuclear power
plants, the transportation of radioactive materials, and natural disasters A sample of
planning aids that address hazardous materials include:
• Hazardous Materials Contingency Planning Course (student manuals)
• Disaster Planning Guidelines for Fire Chiefs
• Disaster Operations: A Handbook for Local Governments
• Objectives for Local Emergency Management
Publications of the Federal Emergency Management Agency relating to a wide variety
of threats to public health and safety can be obtained by writing:
1-3
-------
Federal Emergency Management Agency
Publications Office
500 C Street, S.W,
Washington, D C. 20472
Publications of the U.S. Department of Transportation
The U.S Department of Transportation (DOT) has sponsored a large number of
research studies and demonstration projects related to planning for transportation emergen-
cies involving hazardous materials over the years Appendix E of NRT-1 contains a fairly
comprehensive list of resulting reports A representative sample of current and past available
titles includes:
• Community Teamwork: Working Together to Promote Hazardous Mate-
rials Transportation Safety -A Guide for Local Officials
• A Community Model for Handling Hazardous Materials Transportation
Emergencies
• Risk Assessment Users Manual for Small Communities and Rural Areas
• Manual for Small Towns and Rural Areas to Develop a Hazardous
Materials Emergency Plan; with an Example Application of the Method-
ology in Developing a Generalized Emergency Plan for Riley County,
Kansas
• Community Model for Handling Hazardous Material Transportation
Emergencies: Executive Summaries
• Hazardous Materials Demonstration Project Report: Puget Sound Region
• Hazardous Materials Hazard Analysis: Portland, Oregon
• Hazardous Materials Management System: A Guide for Local Emergen-
cy Managers
• Lessons Learned: A Report on the Lessons Learned from State and
Local Experiences in Accident Prevention and Response Planning for
Hazardous Materials Transportation
The Community Teamwork document may be obtained by writing to*
1-4
-------
Office of Hazardous Materials Transportation
Attention: DHM-50
Research and Special Programs Administration
Department of Transportation
400 7th Street, S.W
Washington, D C 20590
Information on the availability of the Hazardous Materials Management System Guide
and other documents developed for the Portland, Oregon area can be obtained by writing.
Multmomah County Emergency Management
12240 NE Ghzan
Portland, Oregon 97230
Most of the other publications and documents of a similar nature are available from the
National Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161
(telephone 703-487-4650).
Publications of the U.S. Environmental Protection Agency
The EPA has published a series of documents to assist emergency planning personnel
Available titles include
• Introduction to Exercises in Chemical Emergency Preparedness Pro-
grams
• A Guide to Planning and Conducting Table-Top Exercises
A Guide to Planning and Conducting Field Simulation Exercises
• Report of a Conference on Risk Communication and Environmental
Management
• Identifying Environmental Computer Systems for Planning Purposes
• Chemicals in Your Community
These documents may be obtained by writing
Environmental Protection Agency
OS-120
401 M Street, S W.
Washington, DC 20460
1-5
-------
Publications of the Chemical Manufacturers Association
Even before SARA required the assignment of individual facility emergency coordina-
tors to Local Emergency Planning Committees (LEPC's), the Chemical Manufacturers
Association (CMA) established a Community Awareness and Emergency Response (CAER)
program to encourage local chemical plant managers to take the initiative in cooperating with
local communities in the development of integrated emergency plans for response to
hazardous material incidents. The NRT guidance document cited above notes that knowl-
edgeable chemical industry representatives can be especially helpful during the planning
process and advises community planners to seek out local CMA/CAER program participants
More specifically, the document points out that many chemical plant officials are both
willing and able to share equipment and personnel during emergency response operations
The CMA publishes three documents that could prove considerably useful during the
overall planning process, including.
• Community Awareness and Emergency Response Program Handbook
• Site Emergency Response Planning
• Community Emergency Response Exercise Program
These publications are available at nominal cost from the CMA Information on
specific items can be obtained by calling (202) 887-1100 or writing.
Publications Fulfillment
Chemical Manufacturers Association
2501M Street, N.W.
Washington, D.C. 20037
Publications oftheAIChE Center for Chemical Process Safety
Established under the auspices of the American Institute of Chemical Engineers
(AIChE), this being the primary professional society of chemical engineers in the United
States, the Center for Chemical Process Safety has undertaken an ambitious program to
promote and ensure safety at chemical plants. Initial efforts have involved the development
and publication of a senes of safety guideline documents. The first four titles below are
complete and currently available to the public. The latter tides are expected to be published
during 1989 or shortly thereafter.
1-6
-------
• Guidelines for Hazard Evaluation Procedures
Guidelines for Safe Storage and Handling of High Toxic Hazard
Materials
• Guidelines for Use of Vapor Cloud Dispersion Models
• Guidelines for Vapor Release Mitigation
• Guidelines for Chemical Process Quantitative Risk Assessment
• Guidelines for Technical Management of Chemical Process Safety
• Guidelines for Obtaining Process Equipment Reliability Data
• Guidelines for Human Reliability in Process Safety
• Guidelines for Process Control Safety
• Guidelines for Processing and Handling Reactive Chemicals
Information on these and other AIChE publications is available from4
AIChE Publication Sales Department
345 East 47 Street
New York, NY 10017
Other Pertinent Publications
Besides the above fairly recent and generalized planning guides published by the
federal government or industry trade associations, there are several other sources of general
information and data available that may be helpful during the overall emergency planning
process. Selected publications are listed and described in Chapter 14. Publications devoted
to specific topics of possible interest to readers are referenced at appropriate locations
throughout the chapters and appendices that follow.
1-7
-------
-------
2.0 KEY PROPERTIES OF CHEMICAL SUBSTANCES
SOLID
LIQUID
GAS
2.1 STATES OF MATTER
Most materials can exist in more than one physical state, a common example being
ordinary water. It is well known that liquid water will freeze and become a solid at 32
degrees Fahrenheit (°F) at normal atmospheric pressure The temperature of 32°F is known
as the freezing point for this substance Alternatively, this temperature can be referred to as
its melting point For water, both the freezing point and melting point are exactly the same
and well-defined This is true for most other substances, but there are exceptions to this
general rule
At 212°F, liquid water begins to boil at normal atmospheric pressure as it begins a
transition or phase change from a liquid state to a vapor or gas. The specific temperature at
which a liquid boils under a given set of environmental conditions is known as its boiling
point temperature or boiling point for short If the boiling takes place at normal atmospheric
pressure, the more appropriate and accurate phrase to use is normal boiling point or boiling
point at one atmosphere. The importance of qualifying the term boiling point with the words
"normal" and "one atmosphere" will be discussed a bit later For now, it is simply adequate
to note that a great many materials in the environment have their own unique freezing/melt-
ing and normal boiling points which can be radically different than those of water For
example, the petroleum product known as butane, the flammable substance in most
disposable lighters, has a normal boiling point of 31°F and will boil and rapidly vaporize if
spilled as a liquid on a block of ice having a temperature of 32°F A temperature of -216°F
-------
would be required to solidify or freeze the butane to a solid, yet even this very low
temperature would be insufficient to prevent boiling of such substances as liquid hydrogen,
helium, nitrogen, and several others.
Not all substances, incidentally, can exist in all three states of matter in the natural
environment. Some solids undergo a piocess called sublimation upon heating whereby the
solid state directly transforms to a gaseous state without first becoming a liquid A good
example is solid carbon dioxide, also known as "dry ice " Carbon dioxide can only become a
liquid in confinement under special conditions of storage.
2.2 DEFINITIONS OF TEMPERATURE AND HEAT
The discussion so far has demonstrated that the temperature of a substance can
influence its form and properties. There is a great deal more to be said on the subject,
however, so there is value in formal definition of important terms before proceeding We
start with the concept of temperature and the flow of heat and energy from one body to
another.
The dictionary defines the temperature of a substance as its "degree of hotness or
coldness measured on a definite scale " The key word here is scale. In the United States, the
scale with which we are most familiar is the Fahrenheit scale, and most of us are aware that
most of the world uses the Celsius or Centigrade scale, this being a part of the metric system
Both of these temperature measurement systems are considered relative scales because key
numbers are essentially the freezing point and boiling point of water at normal atmospheric
pressure. These numbers are 32°F and 212°F respectively on the Fahrenheit scale and 0°C
and 100°C on the Celsius scale In order to convert from one scale to another, one of two
common equations is used, these being:
degrees F - (1.8 x degrees C) + 32
degrees C = (degrees F - 32)/1.8
It is also useful to know that a one degree change on the Celsius scale is equal to a 1 8°
change on the Fahrenheit scale. Thus, a temperature rise of 18° on the Fahrenheit scale is
equivalent to a rise of 10° on the Celsius scale
Besides these two scales, there are two others that are commonly used in the scientific
community and which are defined as absolute scales in the sense that zero degrees refers to
an absolute lack of heat in the object being measured Absolute zero is about 460° below
zero on the Fahrenheit scale and about 273° below zero on the Celsius scale
2-2
-------
One of these absolute scales is known as the Rankine scale and is related to the
Fahrenheit scale such that a temperature in degrees Rankine equals the temperature in
degrees F plus 460. Thus, 100°F equals 560°R, where the R denotes use of the Rankine
scale.
The second absolute scale is the Kelvin scale and is in very common use by today's
engineers and scientists on a worldwide basis It is related to the Celsius scale such that a
temperature in degrees Kelvin equals the temperature expressed in degrees Celsius plus
273.15. Thus, 100°C equals 373.15°K, where the K denotes use of the Kelvin scale
As noted before, all temperature scales are used to measure and represent the degree of
hotness or coldness of a substance. In actuality, however, this is a somewhat misleading
statement, because heat can be technically defined as "energy whose interchange between a
system and its surroundings takes place only by virtue of a temperature difference " Thus,
heat is a form of energy that increases the temperature of substances and which can flow
from a warm body to one which is cooler Whenever a cold body is placed in a warm
environment, there will be a temperature difference, and heat will flow from the wanner
environment to the colder body. Alternatively, if the body is warmer than its surroundings, it
will lose heat Thus, when a cold liquid is spilled into a warm environment, it will
experience a heat gain. Depending on the temperatures involved, this temperature may be
sufficient to cause the liquid to boil (remember the boiling butane on the block of ice'')
Alternatively, if the liquid was hot to begin with, it may lose sufficient heat to solidify or
freeze The importance of these concepts will become apparent as the discussion turns to the
topic of how a chemical may behave when released into varying environmental conditions
2.3 DEFINITION OF PRESSURE
The next concept to be discussed is that of pressure, which can be defined as the
amount of force brought to bear on some unit area of an object. When we press our thumb
down on a table, we are applying force on the table. The harder we press, the greater the
force, and the greater the pressure we apply to the table surface.
As we sit here, the air in the sky above us presses down on our bodies and all objects
around us with the pressure of approximately 14.7 pounds per square inch of surface area,
commonly abbreviated as 14 7 psi. This pressure, essentially the average air pressure at sea
level, is also known as one standard atmosphere. When one speaks of a pressure of two
atmospheres as might be found in a tank, pipeline, or other container of a hazardous material,
it generally means that 29 4 psi is present, or two times 14 7 psi
2-3
-------
The word generally is emphasized because pressure also has absolute and relative
scales of measurement The 14 7 psi of atmospheric pressure at sea level is an absolute
measurement and is more properly presented in units of pounds per square inch - absolute, or
psia for short. Zero psia in this case refers to a complete absence of pressure such as one
might find in the perfect vacuum of outer space The most common relative scale of
measurement, this being one only used in the United States for the most part, presents
numerical values in terms of gauge pressure, where a reading of zero matches an absolute
pressure of one standard atmosphere. In this system, an absolute pressure of 15.7 psia would
be expressed as 1 0 pound per square inch - gauge, or 1 0 psig for short Two atmospheres of
absolute pressure would be equivalent to one atmosphere gauge pressure
There are several other systems of pressure measurement that are of an absolute nature
The most common include:
• Millimeters of mercury (mm Hg) - in which 760 mm Hg are equivalent to
one standard atmosphere
• Newtons per square meter (N/m2) - in which 101,325 N/m2 are equal to one
standard atmosphere
• Pascals (Pa) - which are another name for N/m2, such that 101,325 Pa are
equal to one standard atmosphere.
• Bars - in which 1 01325 bars are equal to one standard atmosphere
• Inches of water (in H^O) - in which 407 6 in Hp are equal to one standard
atmosphere
• Inches of mercury (in Hg) - in which 29 9 in Hg are equal to one standard
atmosphere
The latter two sets of units are not in as common use in the scientific community as the
first four but it is well to know of their existence Those of you who pay attention to weather
forecasts will recognize that meteorologists have traditionally reported current atmospheric
pressures in units of inches of mercury
2.4 VAPOR PRESSURES OF LIQUIDS AND SOLIDS
Liquids have a tendency to evaporate even at temperatures well below their boiling
points. The reason for this stems from the observation that molecules of a liquid (these being
the smallest parts of the liquid which retain the identity of the substance at the atomic level)
have a tendency to break away from the surface of a liquid and enter the vapor state The
2-4
-------
speed of this process, in the absence of wind effects, is a function of temperature such that a
warm hquid will evaporate more quickly than the same liquid at a cooler temperature Note,
however, that different liquids at the same temperature will evaporate at different rates
depending on their particular properties
One primary measure of a liquid's tendency to vaponze is known as its vapor pressure,
this being the pressure exerted by its vapors on the walls of a container which is partially full
of the liquid and free of any other vapor or gas. Higher temperatures cause increases in the
vapor pressure Lower temperatures cause a decrease, and there is a direct relationship
between the temperature of any given substance and its vapor pressure Table 2 1 provides a
list of vapor pressures for variety of common substances showing how they differ with
respect to temperature Note that the pressures are expressed in units of millimeters of
mercury (mm Hg), this being the most common set of units used for this purpose m the
United States, particularly for substances at temperatures below then- normal boiling points
Note also that there are wide variations in the temperatures associated with specific vapor
pressures and that even iron will have a measurable vapor pressures if heated to very high
temperatures.
The substances listed in Table 2.1, and all others, exert their specific vapor pressures
whether or not they are enclosed in a sealed container When m a container, they reach a
state of equilibrium such that some molecules go from the liquid state to the vapor state
while others pass back from the vapor to the hquid at the same rate, and no material is lost to
the outside environment. When in the open, molecules entering the vapor state mix with air
and move further and further away from the liquid surface with time. As they are replaced
above the surface with new molecules evaporating from the liquid, the volume of liquid is
depleted. Eventually, all the liquid evaporates (be it in minutes, hours, days, or years) and
the surface becomes dry.
Figure 2 1 illustrates these various phenomena. In the top diagram, we observe
molecules evaporating from a pool of hquid and entering the atmosphere Note that any type
of wind or breeze blowing across the surface of the hquid would help the individual
molecules in escaping and moving away from the liquid and thereby increase the overall rate
of evaporation. This rate is indeed a partial function of air velocity over the surface such that
higher velocities usually produce higher evaporation rates
In the middle diagram of Figure 2.1, the liquid is confined within a container and the
escaping vapor molecules are trapped Eventually, as illustrated m the bottom diagram, a
state of equilibrium is attained.
2-5
-------
TABLE 2.1
VAPOR PRESSURES AS A FUNCTION OF TEMPERATURE
Chemical
Benzene
Butane
Ethyl alcohol
Ethylene glycol
Iron
Methyl alcohol
Propane
Water
Vapor Pressure (mm Hg)
1
10
40
100
400
760
Temperature (°F)
-382
-1507
-24.3
127.4
3249
-47.2
-2000
-18*
11.3
-1080
279
1978
3702
28
-163.3
523
45.7
-74.4
662
248.0
4035
41.0
-1343
93.3
79.0
-47.6
948
2872
4280
702
-1113
122.3
1411
27
1463
3533
4721
1218
-681
1765
176.2
31.1
173.1
3871
4955
1485
-43.8
2120
t-o
*Approximate
-------
FIGURE 2.1
EVAPORATION AND VAPOR-LIQUID EQUILIBRIUM PHENOMENA
EVAPORATION MOLECULES ESCAPING FROM
THE LIQUID TO BECOME VAPOR
\ x
f \
Vtf
\r========-)
EVAPORATION MOLECULES CONFINED
VAPOR-LIQUID EQUILIBRIUM
MOLECULES ESCAPE FROM AND
RETURN TO THE LIQUID
2-7
-------
It should be realized that there is a direct relationship between the vapor pressure of an
evaporating substance and the maximum concentration that its vapor or gas may achieve
when mixed with air in the open environment This is true because higher vapor pressures
above the surface of a substance require that more molecules of the substance be physically
present Thus, if the vapor pressure of the substance is known, one can compute the
approximate maximum airborne contaminant (i e, chemical) concentration it may attain.
Such concentrations are most commonly expressed in units of percent in air by volume, parts
per million parts of air (ppm) by volume, parts per billion parts of air (ppb) by volume, or
milligrams of chemical per cubic meter (mg/m3) of air. The equations needed for these
computations are'
^ . vapor pressure (mm Hg) ,._
% concentration = F F ,„ v — x 100
7oU
vapor pressure (mm Hg) ^, ^ nnn
ppm concentration=—-—-—=77: x 1,000,000
ppb concentration = concentration in ppmxlOOO
me (ppm concentration) x (molecular weight)
-concentrate 0 08205 x (273 15+* C)
A restriction to remember in using these equations is that the concentration of a gas or
vapor cannot under any circumstances exceed 100% by volume or its equivalent of 1,000,000
ppm regardless of the answer obtained An example should help the understanding of these
relationships.
From Table 2.1, we find that benzene has a vapor pressure of 100 mm Hg at a
temperature of 79 0°F. From earlier discussion, we also know that 79.0°F is equal to 26 1°C
Therefore:
„ . 100x100 *~~M. ,
% concentration = —=77:— = 13 16% by volume
100x1,000,000 .„„„ . ,
ppm concentration= — = 131,600ppm by volume
760
Computation of the equivalent concentration in mg/m3 requires not only knowledge of
the temperature in degrees Celsius but of the molecular weight (m w) of the material, this
being an atomic measure of the weight of the substance. This weight is often (but not
always) listed in material safety data sheets (MSDS) and product bulletins that present data
2-8
-------
on the physical and chemical properties of chemicals Section 2.8 of this chapter demon-
strates how to compute the molecular weight of a substance given knowledge of its chemical
formula The molecular weight of benzene is 78. 1, so
2.5 BOILING POINTS AS A FUNCTION OF PRESSURE
It was reported earlier that a pressure of 760 mm Hg is equal to 14 7 psia or one
standard atmosphere From Table 2 1 we see that water has a vapor pressure of 760 mm Hg
at a temperature of 212°F, a temperature we recognize as its normal boiling point. What is
significant about this observation is that it holds true for all liquids Any liquid will begin to
boil at the temperature at which its vapor pressure equals the pressure being exerted by the
environment onto the surface of the liquid In practical terms, this means that:
• Boiling points of materials are a function of pressure.
• Liquids in sealed containers (with an exception discussed below) will
remain as liquids when heated above their normal boiling points although
then: vapor pressures may far exceed one standard atmosphere pressure
within the container.
• If heating continues and the pressure is not adequately relieved by a safety
device (such as a pressure relief valve), the pressure and temperature within
the tank may eventually nse to the point that some part or all of the
container may burst or rupture, possibly in a violent fashion This may also
occur if the capacity of the safety device is inadequate to prevent an
excessive buildup of pressure
• Materials exposed to environmental pressures below one standard atmo-
sphere will boil at temperatures below then- normal boiling points. Thus,
water will boil at a temperature below 212°F when heated on top of a high
mountain. Water released into a vacuum at any temperature will almost
instantly vaporize.
It is well to realize that many substances with normal boiling points far below normal
ambient temperatures are shipped or stored in commerce as liquids This is most often
achieved by placing the liquid within a strong tank and permitting it to remain in the liquid
state under its own vapor pressure at equilibrium conditions Examples of the most common
of these materials considered hazardous include liquid anhydrous ammonia, ethylene,
chlorine, ethylene oxide, vinyl chloride, and liquefied petroleum gas (LPG) or propane Such
2-9
-------
substances, frequently referred to as compressed liquefied gases, are particularly hazardous
because: 1) leaks may result in rapid venting of gas to the atmosphere; 2) leaks may result
in discharge of liquids that rapidly flash vaporize and/or boil upon exiting their containers;
and 3) the flammable, toxic, or otherwise hazardous gases and vapors evolved may travel
considerable distances downwind before becoming diluted with air below hazardous
concentrations.
Less frequent in transportation but more common at storage and processing sites are
bulk quantities of substances such as chlorine, anhydrous ammonia, or liquefied natural gas
(LNG) which have been liquefied by cooling to low temperatures via the use of refrigeration
systems. Although the vapor pressure of gases liquefied by refrigeration may be close to
ambient pressures within storage vessels, spills into the warmer external environment will
again result in boiling and the evolution of large quantities of potentially hazardous gases and
vapors
The exception to the "rule" that liquids in sealed containers will remain as liquids when
heated above their normal boiling points involves the fact that this is true only so long as the
temperature of the liquid is below what is referred to as its critical temperature The critical
temperature of a substance is the temperature above which it cannot remain in the liquid state
regardless of any increase in pressure. Thus, substances heated above then: critical tempera-
tures are neither liquid nor gaseous, but rather, in a state somewhere in between. Picture
them as very thick vapors.
2.6 DEFINITIONS OF SPECIFIC GRAVITY AND DENSITY
Boiling points, vapor pressures, and melting or freezing points can tell us much about
how a material will initially behave when released into the environment, but more
information is needed to better define actions and behavior This section discusses relative
and absolute measures of the weights of materials, while the next discusses the degree to
which one substance can mix with another
Every solid or liquid in the environment occupies a specific volume of space and has a
certain weight Thus, we may express the weight density of a substance as its weight divided
by its volume. It is well known, for example, that pure water weighs about 62 4 pounds per
cubic foot (lb/ft3) of volume, which is equivalent to 1 0 gram per cubic centimeter (g/cm3) or
1,000 kilograms per cubic meter (kg/m3) in the metric system We also have observed that
different substances have different weights for the same volume. One cubic foot of oil
weighs about 50 pounds. A cubic foot of steel weighs about 487 pounds
2-10
-------
An alternative method of expressing the weight density of a solid or liquid involves use
of a quantity known as the liquid or solid specific gravity Quite simply, this quantity is
determined by dividing the density of a substance by the density of water. Since 62 4 divided
by 62 4 has a value of 1.0, water has a specific gravity of 1 0 and serves as the reference
point for all materials. The liquid specific gravity of a typical oil is 50 divided by 62.4,
giving a value of 0 80 The solid specific gravity of steel is 487 divided by 62 4 and equals
780
As is the case with vapor pressures, both the density and specific gravity of solids and
liquids vary with temperature. Heat causes most (but not all) materials to expand in volume
while cold causes them to shrink. Since the volume changes while the weight remains the
same, the density of a substance generally decreases with heating and increases with cooling.
This explains why most sources of information on the density of chemicals will provide the
temperature at which the value was measured In the case of specific gravities, they may list
both the temperatures of the water and chemical substance used to determine the specific
gravity.
Knowledge of liquid or solid specific gravities is most important when it is desired to
determine how a substance will behave in the presence of water. For example, the fact that
the specific gravity of a typical oil is 0 80 supports the observation that most oils are lighter
than water and have an initial tendency to float The fact that steel's specific gravity is about
7.80 explains why a block of steel will immediately sink in water.
Discussion of vapor or gas specific gravities and densities is more complicated because
these properties are affected by changes in pressure as well as temperature. However, since
we are primarily interested in chemical substances that escape into the natural environment,
since the natural environment has a nominal atmospheric pressure of one atmosphere, and
since any gas or vapor entenng the atmosphere will quickly adjust its volume to achieve a
total pressure of one atmosphere, it is sufficient for the purposes of this text to only consider
specific gravities and densities of gases or vapors at atmospheric pressure.
The discussion begins with the observation that air has a density of 0 0763 lb/ft3 (about
1.22 kg/m3) at a temperature of 60°F and a pressure of one atmosphere As in the case of
other substances, higher temperatures cause a decrease in density and lower temperatures
cause an increase. Similarly, there is a quantity known as the vapor specific gravity or vapor
density which is a ratio of the density of a pure gas or vapor to the density of air. Found in
many data sources, this specific gravity or density (the former term being used rather
interchangeably with the latter) is based on the assumption that an- has a value of 1 0. Thus,
vapors or gases with vapor specific gravities less that 1 0 are presumably lighter than air in
the natural environment while those with values greater than 1.0 are presumably heavier.
2-11
-------
The word presumably is emphasized because the values for vapor specific gravities
found in all too many data sources are frequently misinterpreted by their users, particularly
and specifically in the case of substances below a temperature that permits them to exist as a
pure vapor or gas at a pressure of one atmosphere. This can lead to incorrect conclusions
about the actions of a vapor or gas upon its release to the environment.
The problem has arisen because many sources compute the vapor density of any
substance by a shortcut method involving division of the molecular weight of the substance
by the molecular weight of air, the latter being approximately 28 9 (as the weighted average
for the mixture of gases that comprise air) Thus, since benzene has a molecular weight of
78.1, these sources will report a vapor specific gravity or density value of approximately
2.70, which to many people suggests that the vapors of benzene in the natural environment
are always 2.70 times heavier than air, which is an absolutely untrue assumption The
misinterpretation results in the belief that benzene vapors will always hug the ground over
considerable distances as they spread from the site of a release and may somehow
accumulate and persist in pits, hollows, basements, or other low lying areas
It was earlier explained that benzene has a vapor pressure of 100 mm Hg at a
temperature of 79 0°F and that this vapor pressure translates into a maximum vapor
concentration directly over the liquid surface of approximately 13 16% by volume It
follows that benzene cannot exist as a pure vapor at this temperature in the natural
environment and that it is incorrect to assume that it is a pure vapor when estimating its
vapor density relative to air (which is what is being done when a molecular weight ratio is
computed). Rather, it is necessary to compare the benzene-air mixture density with the
density of pure air to determine whether the vapors generated by the release will be heavier
or lighter than air. This is accomplished in an approximate fashion via the following
procedure:
Step 1- Compute the approximate density pv of pure chemical vapor in lb/ft3 at
temperature T (in °F).
1 3691 x molecular weight
pV= (T+460)
Step 2: Compute the approximate density pa of air in lb/ft3 at ambient
temperature T (in °F).
39.566
pa=-
(T+460)
Step 3: Compute the relative vapor density of the chemical-air mixture
2-12
-------
r> i *. j {Cxpv} + {(100-C)xpa}
Relative vapor density= *-—*•—— -—*—i
v y lOOxpa
Where C = saturated concentration of the chemical vapor in air in percent
by volume.
Benzene has a molecular weight of 78.1 and a maximum vapor concentration (more
precisely referred to as its saturated vapor concentration) of 13 16% in air at 79°F. Use of
these values in the above equation, together with the assumption of an air temperature of
79°F, provides a true relative vapor density value of 1.22 What this means is that the
benzene-air mixture directly above a pool of benzene at the specified temperature is only
1 22 times heavier than air and not the 2.7 times suggested by the vapor density frequently
reported in the literature for this substance Since this mixture will very quickly mix with
additional air as it drifts away from the pool, it will rapidly approach the density of pure air
and behave as if there were little or no difference in its density. In scientific terms, it will
behave as a neutrally buoyant vapor-air mixture.
If the relative vapor density of a substance under prevailing discharge conditions
exceeds 1.5 (as a general rule of thumb), then vapors or gases may indeed behave as
heavier-than-air (or negatively buoyant) mixtures for some distance from the source of
discharge. Conversely, a relative vapor density significantly less than one suggests that a
vapor-air mixture may be lighter than air (or positively buoyant).
In determining or deciding whether any particular gas or vapor will be negatively,
neutrally, or positively buoyant in air, it is also often necessary to consider the circumstances
under which the substance may be released to the atmosphere For example, in situations in
which a compressed liquefied gas is discharged from a container, particularly when in the
liquid state, the resulting vapor cloud or plume may include a considerable amount of fine
liquid droplets. Although the gas or vapor mixture with air may normally be positively or
neutrally buoyant, the presence of these relatively heavy droplets (also referred to as
aerosols) may cause the cloud or plume to behave initially in a negatively buoyant fashion.
2.7 SOLUBILITY IN WATER
All of us have observed that sugar and salt dissolve in water and seem to disappear, that
our favorite alcoholic beverage can be mixed freely with water-based mixers, and that the
"fizz" in containers of soda pop, tonic, or beer is due to carbon dioxide gas that has been
dissolved in the liquid In each case, the solid, liquid, or gas that has dissolved in water is
said to be soluble in water.
2-13
-------
An important concept to understand is that different materials have different degrees of
solubility. At one extreme, there are liquids which are soluble in all proportions with water
and which are also said to be miscible This means that any amount of the substance can be
added to water and at no point in the process will the substance form a separate layer or
phase. At the other extreme, there are substances which do not dissolve in water whatsoever
and which are considered to be insoluble or immiscible. A somewhat extreme example of
the latter case involves stone pebbles in a glass of water. No matter how hard the pebbles are
shaken or stirred, they will not dissolve or form a solution with water, this being the term
used for a mixture of two liquid substances which are mutually soluble.
In between the above extremes are substances which are partially soluble in water. For
example, there is only a certain amount of ordinary table salt that can be dissolved in water
before any new salt added to the solution simply sinks to the bottom and is unable to
dissolve. In the case of table salt, 35.7 grams of salt will dissolve in 100 grams of water at a
temperature of 32°F and this will nse to about 39.8 grams (there are about 454 grams in a
pound) at a temperature near 212°F. And yes, that means that solubility is also a function of
temperature. Generally speaking, hot liquids can dissolve more of a partially soluble liquid
or solid than cold liquids. Alternatively, because of effects involving vapor pressures and
their increase with temperature, cold liquids can generally dissolve more gases and vapors
than hot liquids. Increases in pressure may also increase the solubility of gases in liquids.
2.8 MOLECULAR WEIGHTS OF CHEMICAL SUBSTANCES
There are approximately 89 natural elements in the world that in various combinations
make up all matter that surrounds us. In addition, a number of man-made elements have
been produced under laboratory conditions involving nuclear reactions and many more have
been theorized but never observed. The atoms of all elements have been assigned individual
atomic weights relative to oxygen by scientists. These are listed in Table 2.2 for most
common elements likely to be encountered in normal commerce and use.
Combinations of various atoms called molecules make up the smallest part of any
chemical compound that retains the specific properties of the substance and have specific
molecular weights that can be computed from the number of atoms of each element present
in the compound, as determined by examination of the chemical formula of the substance.
Such formulae are always found in material safety data sheets for pure substances and many
other sources of chemical data. Examples include.
• Hp for water
• CO for carbon dioxide
2
NaCl for sodium chlonde
2-14
-------
TABLE 2.2
ATOMIC WEIGHTS OF SOME COMMON ELEMENTS
Table gives chemical symbol, name, and atomic weight of each element.
Ag
Al
As
B
Ba
Be
Bi
Br
C
Ca
Cd
Ce
Cl
Co
Or
Cs
Silver
Aluminum
Arsenic
Boron
Barium
Beryllium
Bismuch
Bromine
Carbon
Calcium
Cadmium
Cerium
Chlorine
Cobalt
Chromium
Cesium
10787
2698
7492
1081
13734
901
208.98
7991
1201
4008
11240
14012
3545
5893
5200
13291
Cu
F
Fe
Ga
H
Hg
I
Li
K
Mg
Mn
Mo
N
Na
Ni
O
Copper
Fluorine
Iron
Gallium
Hydrogen
Mercury
Iodine
Lithium
Potassium
Magnesium
Manganese
Molybdenum
Nitrogen
Sodium
Nickel
Oxygen
6354
1900
5585
6972
100
20059
12690
694
3910
24.31
5494
9594
1401
22.99
5871
1600
P
Pb
Rb
S
Sb
Se
Si
Sn
Sr
Ta
Ti
U
V
W
Zn
Zr
Phosphorus
Lead
Rubidium
Sulfur
Antimony
Selenium
Silicon
Tin
Strontium
Tantalum
Titanium
Uranium
Vanadium
Tungsten
Zinc
Zirconium
3097
207.19
85.47
3206
12175
78.96
28.09
11869
87.62
18095
47.90
23803
5094
183.85
65.37
91.22
-------
KOH for potassium hydroxide
for methyl hydrazine
for benzene
As noted earlier, knowledge of molecular weights is required for computation of vapor
concentrations in air in some cases, and indeed, knowledge of this weight is mandatory for a
wide variety of calculations involving hazardous materials. Since molecular weights are not
always found on materials safety data sheets, however, it is worthwhile to understand how
they may be computed using the information provided in Table 2.2. This is best accom-
plished by an example.
From the list above we see that methyl hydrazine has a chemical formula of
CHjNHNHj (which may also be shown as CH6N2 in some references). What this means is
that each molecule of this chemical consists of:
• One (1) atom of carbon represented by the symbol "C"
• Two (2) atoms of nitrogen represented by the symbol "N", and
• Six (6) atoms of hydrogen represented by the symbol "H".
From Table 2.2 we find that the atomic weights of carbon, nitrogen, and hydrogen are
respectively 12.01, 14.01, and 1.00. Thus, we can compute the molecular weight of this
substance by multiplying the atomic weight of each of the three elements by the number of
its atoms in the molecule, and then summing the results For methyl hydrazine, the result is:
Molecular weight = (1 x 12.01) + (2 x 14.01) + (6 x 1 00) = 46.03
2-16
-------
3.0 ACTIONS UPON RELEASE TO THE ENVIRONMENT
3.1 PHYSICAL STATE PRIOR TO RELEASE
The first step in determining how a substance will behave upon release to the
environment requires knowledge of the physical state of the material within its storage or
transportation container This in turn requires knowledge of the relationship between the
temperature of the material, its boiling point, and its melting point The possibilities are:
• The temperature of the material in its container is less than its melting
point, in which case the material is a solid in its container. A good example
would be dry table salt in a large drum
• The temperature is greater than the melting point of the material but less
than its normal boiling point, in which case the material is a liquid and the
container contents are approximately at normal atmospheric pressure. An
example is water in a tank at temperatures above freezing. Such liquids,
however, could also consist of substances which are normally solids but
which have been melted and maintained at relatively high temperatures to
keep them liquid. They could also be substances which are normally gases
in the natural environment but which have been liquefied via refrigeration.
• The temperature is greater than the boding point of the material, in which
case the material is a compressed gas (gas under high pressure in a cylinder
or other container) or a liquefied compressed gas (a substance that is
normally a gas at normal ambient conditions but which has been turned into
a liquid by subjecting it to and maintaining it at high pressures, thus raising
its actual boiling point).
Table 3 1 summarizes the various possibilities in greater detail The table requires a bit
of study for complete understanding, but the effort is extremely worthwhile
32 MATERIAL STATES DURING AND INITIALLY AFTER RELEASE
Once there is an understanding of the state of a hazardous material within a storage or
transportation container, it is next necessary to consider how the substance will behave
initially when discharged into an environment of normal ambient temperatures and pressures
There are 10 scenarios to consider based on the last column of Table 3 1, all of which assume
that the spill or discharge does not take place during a fire or other abnormal event which
would change internal and/or external temperatures
-------
TABLE 3-1
TYPICAL STATES OF MATERIALS IN STORAGE OR TRANSPORTATION CONTAINERS
Normal Melting or Boiling Points
Container Conditions
State of Material (Scenario #)
Melting point less than ambient T
T less than melting point and less
than ambient T
Cold solid (1)
Melting point greater than ambient T
T near ambient T
Solid near ambient T (2)
Boiling point greater than ambient T
T greater than melting point, greater
than ambient T, but less than boiling
point
Warm or hot liquid
(molten solid) (3)
Melting point less than ambient T
T greater than melting point but less
than ambient T and boiling point
Cold liquid (4,5)
N>
Boiling point greater than ambient T
T near ambient T
T greater than ambient T but less
than boiling point
T greater than boiling point and
greater than ambient T
Liquid at ambient T (6)
Hot liquid (7)
Hot or warm compressed gas or vapor over
hot liquid (8)
Boiling point less than ambient T
T near ambient T
T greater than boiling point and
greater than ambient T
Compressed gas or compressed liquefied
gas under pressure at ambient T (9,10)
Hot or warm compressed gas or com-
pressed liquefied gas under pressure at T
greater than ambient (9,10)
Notes: T = temperature within container, ambient T = temperature outdoors
-------
Scenario #1: Cold or Refrigerated Solids
Some materials that are normally liquids or gases at ordinary temperatures or pressures
are handled as solids at temperatures below their melting points and below ambient
temperatures to make them easier or safer to transport or use. When exposed to a warmer
environment, they will melt to become liquids, or if they are substances that pass directly
from a solid to gaseous phase (ie., substances that "sublime") they will vaporize. For
example, ice spilled on the ground in summer will melt to become liquid water. Solid carbon
dioxide (dry ice) will "sublime" as it warms to become carbon dioxide gas.
Scenario #2: Normally Solid Materials
Materials that are solids at ordinary ambient temperatures and which are transported or
otherwise handled at such temperatures will remain as solids upon release from their
containers. Dry table salt and sugar are good examples.
Scenario #3: Molten Solids
Some substances which are normally solids are melted to become liquids, since liquids
are sometimes easier to handle. Indeed, for transportation, a solid may be melted and poured
into a tank vehicle of some kind where it will slowly cool with time, and even possibly
resolidify. When it reaches its destination, it will be pumped out if still a liquid, or first
remelted (possibly using heating coils inside the tank) and then pumped out. Such
substances will either be discharged as solids or as liquids that may solidify if exposed to
cooler ambient temperatures during an accidental spill or discharge situation.
Scenarios #4 and #5: Cold or Refrigerated Liquids
Liquids that are handled at cold temperatures and/or which are refrigerated may have
normal boiling points that are either below or above ambient temperatures. The latter
substances will simply warm up when released to the environment (Scenario #4), much as
cold water will heat in the sun. Those with below ambient boiling points (Scenario #5),
which are typically cooled to reduce their vapor pressures in equipment or for use in
air-conditioning or refrigeration systems, will warm to their normal boiling point tempera-
tures upon release and begin to boil Due to thermodynamic cooling effects associated with
liquid evaporation or boiling, these liquids will remain at their normal boiling points. If
spilled onto a surface that is a good heat insulator, the boiling may eventually slow down or
stop, but the quiescent pool that remains will continue to rapidly evaporate This evaporation
process will maintain the remaining liquid near its boiling point as it picks up heat from its
surroundings
3-3
-------
Scenario #6: Normally Liquid Materials
Materials that are normally liquids at ordinary temperatures and pressures and which
are transported or otherwise handled at such temperatures will remain as liquids upon release
to the environment. Good examples would be gasoline or fuel oils pouring from a hole in a
storage or transportation container.
Scenarios #7 and #8: Hot Liquids
There are many cases where a substance that is a liquid at normal ambient temperatures
and pressure might be heated for one purpose or another. Such liquids, if below then- boiling
point (Scenario #7), will cool upon release to the environment and remain as liquids.
However, if they were heated above their boiling points (Scenario #8), then any space above
the liquid in a container will contain gas or vapor at a pressure in excess of one atmosphere.
What happens in the event of an accident or incident in this latter case will depend on what
part of the container is damaged.
• If the container is punctured or otherwise damaged in the space above the
liquid, vapors of the liquid will blow out (i e, vent) from the resulting hole
into the atmosphere and will continue to do so until the liquid cools below
its boiling point For example, picture steam blowing out the stack of an
old-tome steam locomotive.
• If the container is punctured below its liquid surface, the liquid will pour
out of the hole while some amount of its "flashes" to vapor upon release.
The part that remains as liquid will boil briefly and then slowly cool to
ambient temperature while evaporating. As an example, picture a leak on
the face of an automobile radiator with steam, a hot water mist, and hot
water exiting the leak area.
Scenario #P: Compressed Liquefied Gases
Regardless of whether these liquids are at ambient or higher temperatures, they will
typically be in pressure vessels designed to maintain and withstand substantial pressures As
in the prior case, what happens during an accident or incident will depend on what part of the
container is damaged.
• If the container is punctured or otherwise damaged in the space above the
liquid, the gas will typically vent at high velocity from the resulting hole
into the atmosphere, possibly creating some amount of liquid droplets
during the process The velocity is likely to drop with time as boiling
within the tank cools the mass of liquid (the tank surface may actually
3-4
-------
become quite cold and even frost over due to thermodynamic cooling
effects), but such venting of gas may continue for considerable penods of
time (possibly until no more liquid is left in the tank)
• If the container is punctured below its liquid surface, the liquid may
literally jet from the hole (remember the very high pressure apt to be in the
vapor space over the liquid) and potentially large amounts may flash into
gas or vapor. Indeed, depending on the material and the temperatures and
pressures involved, the tank may blow out a large mass of vapor mixed
with small liquid droplets (an aerosol) to the extent that no liquid reaches
the surface beneath the tank If liquid does reach the surface, it will have a
tendency to form a boiling or rapidly evaporating pool.
Scenario #10: Compressed Gases
Gases which are compressed to high pressures in a container or gas cylinder but which
have not been liquefied will vent from any opening in the container at high velocity. As the
gas vents, the pressure in the container will drop and the container and its contents will cool
At some point, when the pressure within the tank drops to standard atmospheric pressure,
venting will cease or drop to a low rate consistent with the amount of heat that enters the
container from its surroundings
3.3 DISCHARGES ONTO LAND
Up to this point, the discussion has essentially focussed on how the boiling point and
melting point of a substance may affect its actions upon release to the environment It is now
time to consider how the density and solubility of the substance impacts on where it will go
once outside and how there are differences to be considered between discharges on land or
water or into the air. The discussion begins with discharges onto land and again considers
the physical states in which spilled substances may reach a land surface
Cold or refrigerated solids with melting points below ambient temperature will either
melt to form a liquid or sublime when spilled onto a land surface Substances that are
normally solid will remain m the solid state, while molten solids may flow for a time as
liquids and eventually solidify as they cool.
Liquids with boiling points above ambient temperatures will remain as liquids and will
generally cool down or heat up as necessary to approach the temperature of the ambient
environment Those with boiling points below ambient temperatures may boil on a land
surface until most of the liquid has volatilized (i e , vaponzed) Alternatively, as the ground
3-5
-------
surface cools beneath a boiling pool that has been confined by natural or man-made barriers,
the boiling may cease and the remaining quiescent pool may simply evaporate at a fairly
steady rate
Gases or vapors may contaminate a land surface if they are soluble in water and either:
1) it is raining; or 2) water sprays are applied by spill response personnel to absorb,
"knockdown" or otherwise accelerate the dispersion of the gas or vapor in air. The
contamination occurs because the water droplets pick up some amounts of the gas or vapor
and then fall to the ground.
Solids of any kind can contaminate the land surface, and are particularly of concern if
they are soluble in water. In such cases, rain or other sources of water will dissolve the solids
and permit them to soak deeper into the ground in a process called percolation Eventually,
the dissolved chemicals may reach the water table (if any) below the land's surface and
contaminate groundwater supplies serving public, private, or industrial water wells Such
contamination may pose a toxic hazard to the people, animals, and plant life that may be
exposed to the soil or which use the contaminated groundwater for drinking, cooking, or crop
irrigation. Similarly, the dissolved chemicals may cause undesired reactions, contamination,
or corrosion of equipment upon entry to industrial processing equipment relying on well
water. The situation for spilled liquids is about the same except that it must be realized that a
liquid need not be soluble in water to percolate into soil or to contaminate groundwater
supplies. Additionally, it must be noted that liquids are more mobile than spilled solids and
do not necessarily require rain or other sources of water to assist in the spreading of
contamination. The rate at which a liquid substance percolates or otherwise penetrates the
ground is, of course, influenced by many factors. Penetration can be rapid in areas of
extremely high permeability including limesinks, caverns, highly fractured rocks, or fractures
widened by solution.
Other concerns associated with discharges of hazardous materials onto land surfaces
are:
• Combustible substances may be ignited and pose a fire or explosion hazard
(see Chapters 4 and 5).
• Hazardous vapors or gases may be liberated into the atmosphere from
substances with significant vapor pressures at prevailing chemical or
ambient temperatures (see Section 3 5 and Chapters 4 and 6).
• Solids, solids dissolved in or earned by land surface water runoff, or liquids
may flow into drains or sewers leading to bodies of water or may directly
contaminate such bodies (see Section 3.4).
3-6
-------
3.4 DISCHARGES INTO WATER
Discharges of a chemical substance into a body of water may occur directly from
damaged ships, barges, underwater pipelines, or railroad cars or trucks that have fallen into
the water, or indirectly, as outflows from sewer or drain outfalls, runoff from spills on land,
runoff of water used to control fires, or entry of contaminated groundwater into the water
body. Virtually all key physical and chemical properties of hazardous materials discussed in
this document play important roles in determining how a material will behave when spilled
into water.
• The boiling point and vapor pressure of the material will determine whether
some part of the material will boil off or otherwise vaporize upon
contacting the water.
• The liquid or solid specific gravity or density of the material will determine
whether it has an initial tendency to float or sink in water.
• The solubility of the material will determine whether it will dissolve in the
water and the rate at which this will occur.
Table 3.2 describes the expected behavior of spills into water of materials with varying
combinations of boiling point, vapor pressure, specific gravity, and solubility attributes To
be stressed is that the table describes spill behavior only m the minutes and hours directly
after a release and that longer penods of time may result in different effects. For example,
although it is well appreciated that oil will float on water, forming a surface slick that may
foul shorelines, it is not as well known that waves, water, turbulence, and time may
eventually cause a floating petroleum oil slick to emulsify (i e., to become tiny droplets) that
distribute themselves through the water column (i.e., throughout the depth of the water
body), to dissolve in water to some extent, and to eventually settle on the bottom of the water
body as a sludge. This sludge, in the case of petroleum oils, may mix with sand or dirt and
form the "tar balls" often observed on shorelines after an offshore spill.
One special point to be made about substances often described as insoluble is that
many of these may actually dissolve at such a slow rate in water that they are considered
insoluble "for all practical purposes" Given enough time or agitation, a sufficient amount
may actually dissolve to cause a toxic hazard to anybody or anything exposed to the
contaminated water Always be wary of claims of complete insolubility when a highly toxic
substance has spilled into water.
3-7
-------
TABLE 3.2
EXPECTED BEHAVIOR OF SPILLS INTO WATER
Boiling Point
Vapor Pressure
Specific Gravity
Solubility
Expected Behavior in Water
Below ambient
Very high
Any
Insoluble
All of the liquid will rapidly boil from the surface of
the water. Underwater spills will most often result in
the liquid boiling and bubbles rising to the surface.
Below ambient
Very high
Less than Water
Low or Partial
Most of the liquid will rapidly boil off but some
portion will dissolve in the water Some of the
dissolved material will evaporate with time from the
water. Underwater spills will result in more dissolu-
tion in water than surface spills.
Below ambient
Very high
Any
High
oo
As much as 50% or more of the liquid may rapidly
boil off the water while the rest dissolves in water
Some of the dissolved material will evaporate with
time from the water. Underwater spills will result in
more dissolution in water than surface spills. Indeed,
little vapors may escape the surface if the discharge
is sufficiently deep
Above ambient
Any
Less than Water
Insoluble
The liquid or solid will float on water. Liquids will
form surface slicks. Substances with significant
vapor pressures will evaporate with time.
Above ambient
Any
Less than Water
Low or Partial
The liquid or solid will float on water as above but
will dissolve over a period of time. Substances with
significant vapor pressures may simultaneously
evaporate with time
Above ambient
Any
Less than Water
High
These materials will rapidly dissolve in water up to
the limit (if any) of their solubility Some evapora-
tion of the chemical may take place from the water
surface with time if its vapor pressure is significant.
-------
TABLE 3.2 (Continued)
EXPECTED BEHAVIOR OF SPILLS INTO WATER
Boiling Point
Vapor Pressure
Specific Gravity
Solubility
Expected Behavior in Water
Above ambient
Any
Near Water
Insoluble
Difficult to assess Since they will not dissolve, and
since specific gravities are close to water, they may
float on or beneath the surface of the water or
disperse as blobs of liquid or solid particles through-
out the water column. Some evaporation of the
chemical may take place from the water surface with
time if its vapor pressure is significant.
Above ambient
Any
Near Water
Low or Partial
Although a material with these properties will be-
have at first like materials described directly above,
it will eventually dissolve in the water Some evapo-
ration of the chemical may take place from the water
surface with time if its vapor pressure is significant.
VO
Above ambient
Any
Any
High
These materials will rapidly dissolve in water up to
the limit (if any) of their solubility. Some evapora-
tion of the chemical may take place from the water
surface with time if its vapor pressure is significant.
Above ambient
Any
Greater than Water
Insoluble
Heavier-than-water insoluble substances will sink to
the bottom and stay there. Liquids may collect in
deep water pockets.
Above ambient
Any
Greater than Water
Low or Partial
These materials will sink to the bottom and then
dissolve over a period of time.
Above ambient
Any
Greater than Water
High
These materials will rapidly dissolve in water up to
the limit (if any) of their solubility. Some evapora-
tion of the chemical may take place from the water
surface with time if its vapor pressure is significant
-------
A second point to be made is that concentrations of water soluble contaminants in
water are typically measured or expressed in units of parts (of contaminant) per million parts
(ppm) of water on a weight basis or in units of milligrams (of contaminant) per liter (mg/l) of
water. These units are essentially equivalent such that one ppm equals one mg/l When a
material is dissolved in water, the mixture is often referred to as an aqueous solution of the
material. Conversely, materials that do not contain water are considered to be anhydrous.
Of interest with respect to the evaporation of chemicals from water is that such
evaporation can take place not only from floating pools or slicks of chemicals but from the
surface of solutions. It is important to realize, however, that the vapor pressure of a chemical
will drop when the chemical is added to water or water is added to the chemical, and the less
chemical there is in the solution, the lower its vapor pressure will be at a specific
temperature. Thus, evaporation from a concentrated solution (i e, one containing consider-
able chemical) near a spill site might create a downwind vapor hazard, but the hazard might
be negligible some time later after the chemical has had a chance to mix with more water.
Similarly, a water-soluble chemical or solution that has a flammable vapor concentration
above its surface at a given temperature may often be rendered nonflammable by the addition
of a sufficient quantity of water.
Besides generating flammable or toxic vapors, chemicals spilled into water or sewers
can pose a variety of hazards to the public and the environment.
• Flammable chemicals or solutions can pose a fire or explosion hazard in
sewers, water treatment facilities, or any other spaces they may enter when
extracted from a body of contaminated water.
• Insoluble materials, particularly oils, may cause drowning of waterfowl
because of loss of buoyancy, exposure due to loss of the insulating capacity
of feathers, and starvation and vulnerability to predators due to lack of
mobility. Coating of the gills of fish may cause death due to lack of
oxygen. Coating of any life forms on the bottom of a water body can kill
by smothering.
• Any insoluble or soluble toxic substance that contaminates water may
poison animals (including humans) or plant life (aquatic plants or irrigated
crops) exposed to the water.
• Organic substances can potentially kill fish and other aquatic life forms by
lowering the oxygen content of the water via biological as well as chemical
processes.
3-10
-------
• Contaminated water drawn into industrial processes may corrode or
otherwise damage or destroy equipment, and possibly cause fire or
explosion hazards.
3.5 FUNDAMENTAL CONCEPTS PERTAINING TO DISCHARGES INTO AIR
Hazardous vapors or gases, ie., those that are flammable or toxic to man or his
environment, may enter the atmosphere from several sources.
• They may be vented directly into the air from a pressure relief valve,
"smokestack", ruptured reaction vessel, broken pipe, or other item of
equipment at a chemical plant or other fixed site facility.
• They may be vented directly from a pressure relief valve, broken valve,
loose fitting, or puncture in a transportation vehicle, container, or cylinder.
• They may evolve from volatile liquids or solids discharged onto the ground
or into water.
Evaluation of vapor or gas discharge hazards first requires that the duration over which
the discharge takes place be characterized It then requires assessment of how the liberated
vapors or gases will mix with air over time in a process referred to as vapor dispersion, and
finally, requires knowledge of the specific hazards posed by exposure of people to resulting
concentrations of airborne contaminants at downwind locations.
Instantaneous vs. Continuous vs. Finite Duration Discharges
The most common methods available for assessment of vapor dispersion hazards
require that discharges of vapor or gases into the atmosphere be classified as either being
instantaneous or continuous in duration. Instantaneous discharges are those that take place
over the course of a few seconds or a minute or so and then stop for all intents and purposes
The result of such a discharge is typically a puff of vapor or gas or a distinct cloud
Continuous discharges take place over longer periods of time and produce long stretched-out
plumes of gas or vapor such as those typically seen from continuously operating
smokestacks. These cases represent the two extremes by which contaminant emissions may
be characterized In the real world, many discharges may be of too long a duration to be
characterized as truly instantaneous, yet too short in duration to establish a continuous plume
These latter discharges are commonly said to be of finite duration,
In the following, we concentrate upon describing the behavior of gases or vapors
liberated from instantaneous or continuous discharges to the atmosphere, thus providing the
reader an understanding of the two possible extreme cases Realize, however, that the actual
3-11
-------
behavior of a volume of contaminated air dispersing in the atmosphere, particularly if
generated from a finite duration discharge, will behave in a manner somewhat between these
two extremes.
Dispersion of Vapor Puffs or Clouds
Picture if you will a large semi-spherical puff or cloud of a pure chemical vapor or gas
on the ground that has somehow entered the atmosphere over the period of a few seconds and
has a vapor specific gravity and vapor density similar to that of air. As the wind pushes on
the puff, the puff will begin to move in the direction of the wind at a similar velocity
Simultaneously, air will begin to mix with the surface of the puff, thus diluting surface
vapors. As more and more air mixes with the puff, the volume of the contaminated airspace
will become larger and larger. Dilution with air, however, will cause vapor concentrations to
drop with time at any point in the puff, although the central core of the puff may remain pure
for a while.
What happens over time and distance as a puff disperses in air is somewhat hard to
visualize with words alone, so it is worthwhile to use various illustrations and graphs in this
endeavor. Figure 3.1 shows four initial stages as a puff moves downwind, each accompanied
below by a graph of vapor concentration in air on the ground along a cross-section of the
puff. At time equals zero, the instant the puff is formed, the concentration within the puff is
close to 100% pure vapor and the air surrounding the puff is uncontaminated At time equals
20, the puff has grown hi size by mixing with air, and that portion which is still 100% pure
vapor has become smaller. The vapor concentration in the remainder of the cloud ranges
from 100% at the edge of the pure core of vapor to 0% at the edge of the cloud By time
equals 40, the core of 100% vapor has become even smaller, and by time equals 80, it has
just disappeared. From this point onward, the peak or maximum groundlevel concentration
will drop below 100% and continue to drop steadily.
Figure 3.2 continues the above sequence for a variety of later times on a single graph.
What is happening is that the cloud grows larger and larger but its peak concentration, the
point at its center along the ground, becomes lower and lower At some point, this peak level
will drop below whatever concentration level of the gas or vapor in an- is considered
hazardous. If one were to plot the groundlevel peak concentration at the center of the cloud
with time or distance, it would resemble the graph in Figure 3.3
Yet another useful way to look at cloud or puff dispersion is to look at the ground area
covered by a particular preselected concentration (which could be a flammabihty or toxicity
limit of some kind). Figure 3.4 demonstrates how this ground area changes from the point at
which the puff is generated to the downwind location that every point in the puff is below the
selected concentration The view is looking down at the puff from a point up above, with
3-12
-------
FIGURE 31
INITIAL STAGES OF VAPOR CLOUD OR PUFF DISPERSION
Wind direction
100% pure vapor
Time = 0
100%—,
100% air out here
Air-vapor mixture
Time = 20
Time = 40
Time = 80
0%
Downwind distance
-------
100%
O
fl
o
0%
Time=80
Figure 3.2
Cross—Section of Puff
Concentration vs Time
Time=160
Time=320
Time=640
Distance Crosswind
100%
a>
o
a
o
o
S
0
s
•»s
X
CO
Figure 3.3
Maximum Puff Concentration
vs. Time or Distance
0%
Time or Distance —
3-14
-------
FIGURE 3 4
PUFF OR CLOUD ISOPLETHS AT INCREASING TIMES
Wind Direction
All circles are
isopleths for the
same concentra-
tion in air at
different times
FIGURE 35
ISOPLETHS IN A CONTINUOUS PLUME
Wind Direction
Larger ovals
represent lower
concentration iso—
pleths All can
exist simultaneously
in a continuous
plume
3-15
-------
each circle representing a different point in time The line around the set of circles encloses
the ground area that will be subjected at some time to airborne contaminant concentrations at
or above the preselected concentration. In somewhat technical terms, the individual circles,
these being lines of constant concentration, are referred to as isopleths, as is the line
enclosing the entire set of circles The latter is also sometimes referred to as the cloud's
footprint on the ground for a particular hazardous concentration.
The downwind distance that any point in puff, cloud, or plume will travel within any
elapsed period of time is related to the velocity of the wind in its direct vicinity by the
relationship:
Distance = Wind Velocity x Time
Although this expression seems rather simple and straightforward, there is a "catch" to its
general use. As observed above, the distance traveled is proportional to the wind velocity in
the direct vicinity of the puff Meterologists and weather stations typically report the
velocity or speed of the wind as it has been measured at a point 10 meters (about 33 feet)
above the ground, where the velocity is usually greater than that very close to a ground
surface. Indeed, volumes of contaminated air released at groundlevel may travel as little as
50 percent of the distance given by the above relationship when the wind velocity used in the
equation is measured at a 10 meter height. Clouds, puffs, or plumes liberated to the
atmosphere above this height may travel faster than the reported wind velocity
Dispersion of Continuous Plumes
As noted previously, the emission of gases or vapors to the atmosphere over an
extended period of time results in establishment of a vapor or gas plume. Points downwind
of the source of emissions will be exposed to a relatively constant airborne contaminant
concentration for a penod of time approximately equal to the duration of the emission so
long as the wind direction holds steady. Note however, as is also the case in instantaneous
discharges, that some amount of time will be necessary for the front edge of a cloud or plume
to reach downwind locations after the initiation of a discharge and for contaminant
concentrations to rise to relatively constant levels at these locations A similar amount of
time will be necessary for the trailing edge to pass downwind points after cessation of vapor
or gas liberation and for contaminant concentrations to drop below levels deemed to be safe.
Thus, there are different arrival and departure times associated with different downwind
locations for both clouds and plumes
Figure 3 5 shows an example of what various concentration isopleths look like through
a horizontal cross-section of an established plume The innermost isopleth encloses the area
3-16
-------
subjected to the highest concentrations. Moving out from the innermost isopleth, each
isopleth in the outer direction represents a lower concentration than the previous isopleth As
in the previous case, the view is looking down from above.
3.6 VARIABLES THAT INFLUENCE ATMOSPHERIC VAPOR DISPERSION
There are numerous factors that influence the size and shape of downwind hazard
zones resulting from vapor or gas discharges into the atmosphere. The most important of
these variables are discussed individually and sometimes in combination below Since
several of them interact with each other, it may be a good idea to read this section more than
once to better understand various interrelationships A solid understanding of vapor cloud
and plume behavior under various conditions is an important prerequisite to proper
emergency response as well as emergency planning
Effect of Toxic or Flammable Limit Selection on Hazard Zone Size
As explained in prior discussions, the concentration of an airborne contaminant
decreases with increasing distance along the downwind centerline direction of the cloud or
plume path as well as in the crosswind direction What this means in practical terms is that
the choice of a higher toxic or flammable limit for definition of hazard zone boundaries
during accident consequence analysis efforts will result in a smaller overall hazard zone than
if a lower limit had been chosen. Conversely, lower limits will lead to larger hazard zones
than higher limits.
The choice of an appropriate toxic limit, also referred to as a "level of concern" in
earlier guidance documents published by the federal government, is discussed in Chapter 6
Flammable limits are discussed in Chapter 4.
Effects of Discharge Rates and Amounts on Vapor Dispersion
In the case of instantaneous discharges and others of relatively short duration, the total
amount (i e., weight) of vapor or gas released to the atmosphere has an impact on the size and
shape of downwind hazard zones All other factors being equal, larger discharge amounts
will result in longer and larger downwind hazard zones. Smaller amounts will result in
shorter and smaller zones
The case with continuous releases is similar All other factors being equal, higher
discharge rates will produce longer and larger hazard zones. Lower discharge rates will
produce shorter and smaller zones
3-17
-------
The area from which a vapor evolves is particularly important when the vapor
originates from a boiling or evaporating pool of liquid A smaller pool will usually evolve a
lesser amount of vapor per unit of time than a larger pool and therefore pose less of a
downwind hazard. A larger pool, having a greater surface area, will produce vapors at a
higher rate and pose a greater downwind hazard Thus, control of exposed pool surfaces can
provide some degree of control over adverse downwind impacts.
Effects of Atmospheric Stability Conditions on Vapor Dispersion
The time of day, the strength of sunlight (if any) in the area, the extent of cloud cover,
and the wind velocity all play major roles in determining the level of turbulence in the
atmosphere and thereby the distances downwind over which airborne contaminants will
remain hazardous Meteorologists typically categorize atmospheric conditions into six
atmospheric stability classes that range generally from "A" to "F". Class A represents
unstable conditions under which there are strong sunlight, clear skies, and high levels of
turbulence in the atmosphere, conditions which promote rapid mixing and dispersal of
airborne contaminants. At the other extreme, atmospheric stability Class F represents light
steady winds, nighttime skies, and low levels of turbulence Airborne contaminants mix and
disperse far less slowly with air under these conditions, which also include atmospheric
inversions (when temperatures increase with altitude rather than decrease as usual), and may
travel much farther downwind at hazardous concentrations than in other cases Table 3 3
denotes the various criteria used for determination of these stability classes Information on
the percentage of time that any particular locale expenences the conditions associated with
each class can be generally obtained from the nearest office of the National Weather Service,
which is listed under the heading of U.S. Department of Commerce in telephone directories
of major cities. Meteorologists associated with local radio and television stations or airports
will also be knowledgeable of these statistics.
During an actual emergency, it will be necessary to understand that atmospheric
conditions may change with time and that these changes will influence the behavior of the
dispersing cloud or plume. As inspection of Table 3.3 reveals, atmospheric stability vanes
strongly with time of day, wind speed, extent of cloud cover, and strength of sunlight As we
are aware, these are all highly variable factors, possibly changing on an hour by hour basis m
some locations during certain seasons
Gas or Vapor Buoyancy Effects on Vapor Dispersion
The descriptions of vapor cloud and plume behavior given earlier started with the
assumption that the vapor specific gravity or density of the gas or vapor being released is
approximately equal to that of air. However, as also discussed earlier, certain gases or vapors
and their initial mixtures with air may actually be heavier or lighter than air.
3-18
-------
TABLE33
ATMOSPHERIC STABILITY CLASS SELECTION TABLE
A ~ Extremely Unstable Conditions
B ~ Moderately Unstable Conditions
C - Slightly Unstable Conditions
D — Neutral Conditions*
E - Slightly Stable Conditions
F — Moderately Stable Conditions
Surface Wind
Speed, mph
Less than 4.5
45-67
6.7-11.2
11.2-134
Greater than 13.4
Daytime Conditions
Strength of sunlight
Strong
A
A-B
B
C
C
Moderate
A-B
B
B-C
C-D
D
Slight
B
C
C
D
D
Nighttime Conditions
Thin Overcast
greater than or =
4/8
Cloudiness**
-
E
D
D
D
less than or = 3/8
Cloudness
-
F
E
D
D
H
VO
*Apphcable to heavy overcast conditions day or night
**Degree of Cloudiness = Fraction of sky above horizon covered by clouds.
-------
In general, lighter-than-air gases, vapors, or mixtures will mix with air in the same
fashion as those that are closer to the vapor specific gravity of air. Groundlevel contaminant
concentrations are likely to be lower, however, because maximum concentrations along the
centerline of the cloud or plume will tend to be elevated The rate at which a cloud or plume
will rise as it moves downwind will primarily be a function of the difference in vapor specific
gravity between it and air and the prevailing wind speed. Lighter gases or vapors will rise
faster. Strong winds will tend to keep the cloud or plume closer to the ground for longer
periods of time. Figure 3.6 for distinct clouds or puffs demonstrates these concepts and the
principles that also apply to plumes In both cases, it is necessary to remember that the
velocity of the wind will influence downwind travel distances within any given period of
time.
Heavier-than-air gases, vapors, or mixtures tend to hug the ground for a time when first
released and may even follow terrain in directions across or against wind directions on
certain boundaries. However, as these vapors and gases become more diluted with air, they
will at some point begin behaving like mixtures with vapor specific gravities close to that of
air. Thus, consideration of heavy gas or vapor dispersion phenomenon is more important for
higher concentrations near the source (such as those associated with lower flammable limits)
than for the lower concentrations typically associated with toxic limits.
The overall behavior of a heavy (negatively buoyant) cloud or plume can be very
different than that of a neutrally or positively buoyant cloud or plume and the shape and
dimensions of the cloud or plume can be strongly influenced by the duration of the discharge,
prevailing atmospheric stability conditions, and prevailing wind velocities. For example, an
instantaneous discharge of a flammable liquefied gas can result m a flammable or potentially
explosive cloud that is 25 percent greater in maximum width than its length under neutral
atmospheric conditions (see Table 3.3) when winds are of moderate velocity Under more
stable atmospheric conditions with lower wind speeds, the maximum width of the cloud
could drop to approximately 80 percent of its length. Under specific combinations of
conditions, particularly for large releases, cloud widths could be as much as 150 percent of
length dimensions.
Continuous or otherwise prolonged discharges of heavy gases or vapors can behave yet
differently from short-term releases Under neutral atmospheric stability conditions, maxi-
mum plume widths typically range from 30 to 60 percent of lengths when winds are of
moderate velocity Under stable conditions, these widths can vary from 75 to 90 percent of
lengths. In contrast, the maximum widths of neutrally or positively buoyant clouds or
plumes are typically in the range of 40 to 50 percent of lengths.
3-20
-------
Figure 3.6
Behavior of Lighter than Air Puffs or Clouds
Wind Direction
Rise of cloud in low wind conditions
Rise of cloud in high wind conditions
Time or Distance
-------
Effects of Source Elevation on Vapor Dispersion
Although many discharges of gases or vapors are likely to take place at or near
groundlevel, some may occur from the top of an elevated item of equipment or from a tall
smokestack, pressure relief valve, or similar venting device. The principles set forth earlier
with respect to post-discharge behavior of gases and vapors remain applicable in such cases,
but it must be noted that groundlevel concentrations due to elevated sources may vary
significantly from groundlevel concentrations due to groundlevel sources. Figure 3.7
illustrates some of the reasons for such differences.
The most important concept to understand about elevated discharges is that maximum
concentrations will be along the centerline path of cloud or plume travel in the downwind
direction. In the case of neutrally buoyant clouds or plumes, groundlevel contaminant
concentrations may be essentially zero until the bottom of the cloud or plume first touches
ground. These concentrations will then rise with increasing downwind distance, reach a
peak, and then drop with further distance. As demonstrated by the graph in Figure 3.3
presented earlier, this differs markedly from the variation of concentration with distance seen
along the centerline path of such a cloud or plume.
When vapors or gases are lighter than air and therefore positively buoyant, the presence
of harmful contaminant concentrations near groundlevel will strongly depend upon the wind
velocity. As illustrated in Figure 3.6, the cloud or plume may nse quickly, slowly, and
possibly not all depending on the wind speed (and the velocity with which the vapors or
gases are discharged upwards into the air). Groundlevel concentrations will vary according-
ly.
Effects on Dispersion Relating to Physical States of Contaminants
Although the discussion to this point has focused on the dispersion of gases and vapors
in air, it is also important to understand that fine mists, fumes, or aerosols of liquids as well
as fine dusts or powders may also be transported by the wind to downwind locations. Some
discharges could involve mixtures of chemical vapors and aerosols and dusts.
Larger and heavier droplets of liquid or particles of solids may "settle out" of the cloud
or plume and drop to ground surfaces fairly close to their point of origin Somewhat smaller
particles may settle out a bit further downwind, while the smallest of all may travel as far as
vapors and gases at equivalent concentrations in air. Droplets of volatile liquids may
vaporize as they are earned by the wind or after they settle out of the main cloud or plume.
They may also cause part or all of a cloud or plume to behave as if it is heavier than air even
if the same substance in a purely gaseous state might be lighter than air or neutrally buoyant
at prevailing temperatures. All of these phenomena can have an impact on groundlevel or
close to groundlevel contaminant concentrations, generally resulting in levels above those
3-22
-------
Wind Direction
1 Puff dispersion of neutrally buoyant vapors Groundlevel
concentrations may be zero for some time until the puff first
hits" the ground Same puff shown at different times above
2 Continuous plume dispersion of neutrally buoyant vapors in
air Note again that some distance may be required before any
contamination occurs near the ground
3 Plume dispersion of heavy vapor Puffs may follow a similar
path during dilution with air
Figure 3.7
Some Effects of Elevated Emissions
3-23
-------
that would be expected in the absence of mists, fumes, aerosols, or dusts. Accurate
prediction of cloud or plume behavior under these conditions is extremely complex and
prone to substantial errors.
Effects of Discharge Velocities on Dispersion
Vapors or gases may be released to the atmosphere at relatively low velocities or may
be vented under high pressure as a jet There are various "jet" momentum effects that alter
puff or plume behavior, particularly near the source of a discharge A strong jet of vapor or
gas will tend to entrain and mix with air rapidly at first, thus tending to reduce contaminant
concentrations These effects become less important, however, as the puff or plume moves
downwind.
In the event that a high velocity high rate discharge of a heavies? than air mixture of gas
and liquid aerosols takes place in the downwind direction, there is a distinct possibility that
downwind hazard zone lengths will be greater than those predicted by most vapor dispersion
models in general use The behavior of such highly pressurized discharges of compressed
liquefied gases is a subject receiving considerable attention in scientific cicles at present, but
accurate prediction of contaminant behavior under these conditions remains prone to
substantial errors
Effects of Local Terrain on Vapor Dispersion
In virtually all that has been said about atmospheric vapor dispersion phenomena up to
this point, it has been tacitly assumed that the vapors or gases being discussed are dispersing
over flat terrain without obstacles of any kind In the real world, however, large portions of
the country are by no means flat or devoid of hills, mountains, trees, or buildings All of
these topographical features and others influence the manner in which airborne contaminants
disperse
In most cases, a certain degree of "roughness" in the terrain is beneficial in the sense
that it tends to speed up the rate at which contaminants mix with air and are thereby diluted
This is understandable if one thinks about how the wind behaves as it swirls around and over
trees, hills, buildings, and other objects There are two situations, however, in which terrain
effects may cause increased hazards at or near groundlevel locations
The first case involves situations in which contaminants are trapped within some sort of
canyon, valley, or bowl-like depression in the land surface Under these conditions, the walls
or sides of these topographical features can prevent spreading of clouds or plumes and
restrict dilution with air. The net result is that hazard zones might be of different size and
3-24
-------
shape than otherwise expected If an atmospheric inversion were to occur such that there was
essentially a "cap" placed over a bowl-like depression or valley, airborne contaminants could
be literally trapped for extended periods of time
The second case involves the dispersion of gases or vapors from an elevated source
when there are buildings or similar shaped features on the land in the downwind direction.
As the wind passes over a building, some part of it may be drawn down into a swirling eddy
pattern in a space behind the structure commonly referred to as its "wake cavity". The
practical significance of this phenomenon is that contaminants liberated from elevated
sources could potentially be drawn down towards groundlevel much sooner and at much
shorter downwind distances than might otherwise be expected
Readers should be advised that many of these phenomena are extremely difficult if not
impossible to address in any sort of generalized vapor dispersion hazard prediction model or
methodology regardless of its claimed level of sophistication or cost Those who may be
tempted to purchase any expensive software package to evaluate downwind vapor dispersion
resulting from chemical accidents for planning purposes should first compare the results of
the package with the results obtained from the computer program provided with this guide
for several scenarios
Effects of Wind Meandering on Evacuation or Protective Action Zones
The main reason that one would wish to determine or predict the concentration
isopleths or footprints of gas or vapor clouds or plumes is to determine those downwind areas
that may require public evacuation or other protective action in the event of a toxic and/or
flammable vapor or gas release. It is important to realize, however, that the direction of the
wind is rarely steady over any significant penod of time and that the wind direction tends to
shift back and forth between various directions. This shifting over time is often referred to as
meandering. The practical significance of wind meandering is that an area larger than that
predicted by strict application of dispersion estimation methods may require evacuation or
other means of public protection during an actual emergency.
The probability and extent of wind meandering in any locale is a complex function of
several factors, but one of the most important involves the atmospheric stability class
prevalent in the area at the time The wind tends to meander less on average under stable
conditions than in unstable weather
Based on data presented on page 28 of the Handbook of Atmospheric Diffusion (U S
Department of Energy, DOE/TIC-11223) by Hanna, Bnggs, and Hosker, it has been
determined that there is a 90 percent probability on average that a cloud or plume will remain
within a downwind arc of 120 degrees from its point of origin under atmosphenc stability
class conditions A, B, and C For more stable stability classes, the arc narrows to a 40 degree
3-25
-------
angle. Figure 3.8 illustrates these observations, the practical significance of which is the
finding that the area requiring evacuation or other protective action (such as sheltenng
populations in place) in the first hour of a hazardous vapor or gas release should usually be
based on the above arcs and not the actual width of selected concentration isopleths Where
an evacuation is to be attempted, it is often best to start from a point nearest the emission
source and work outwards towards downwind areas subject to lower concentrations Be
advised, however, that there is an exception to the above findings When the velocity of the
wind is very low under special circumstances, the direction of the wind can become very
erratic. It is best to be prepared under low wind conditions for one or more sudden shifts in
wind directionW the possibility that a cloud or plume may literally "hop" from one position
or direction to another
Indications of the specific areas that may require protective action in the event of
specific spill or discharge situations can be obtained by drawing hazard zone boundaries on a
map of the region in accordance with the "scale" shown on the map These boundaries can
be drawn for various wind directions and atmospheric stability classes to illustrate potential
hazard zones under various conditions Local census data may then be used to estimate the
maximum number of people that may require protection Note that the drawings will be most
easily drawn using a ruler and protractor, keeping in mind that a full circle has 360 degrees.
If the discharge or release may be prolonged, the probability will increase that there
will be a major shift in wind direction When and where possible, it is best to consult a
meteorologist with detailed knowledge of current local conditions immediately for advice on
how to expand the evacuation area as time progresses For truly prolonged situations
involving hazardous emissions, it may eventually become necessary to evacuate a full circle
around the accident site out to the radial limits of the estimated hazard zone
It is precisely because the direction of the wind during an accident cannot be predicted
in advance and that the direction may shift during a hazardous event that the zone considered
vulnerable around a potential accident or incident site encompasses a full circle around the
site (or a "corridor" of overlapping circles if the site is along a railroad, pipeline, barge, or
truck route) Although there may be many cases in which only a portion of the vulnerable
zone will require protective action, public and industry officials must realize that the entire
zone is at risk and will require attention during the emergency planning process, particularly
with respect to populations at special risk or requiring special assistance
The average probability of the wind being in any particular direction may be useful
knowledge, particularly in locations where the wind is prone to flow in certain directions on a
regular basis during various seasons As in the case of atmospheric stability classes, the
planning process can therefore benefit from consultation with a meteorologist at the nearest
office of the National Weather Service 01 associated with a local radio or television station or
3-26
-------
FIGURE 3.8
VAPOR DISPERSION HAZARD ZONE BOUNDARIES
Wind direction
t
40 degree arc
Use for discharges less
than one hour in duration
under atmospheric stability
conditions D, E, and F
\
, \ \ \ X
I—i—V—«- \ -N:
120\ degre^ a\c
Uac \for—discharges
less \than one Sour
undeb- atmospher
stability conditio
A, B,\or C
Emission source
Arrows within boundaries of estimated hazard zones
indicate length of downwind hazard distance in downwind
centerline direction of wind
Longer duration discharges may require up to 360 degree
evacuations or protective measures if the wind direction
may shift during the discharge. Consult a qualified
meteorologist in actual emergencies for advice.
3-27
-------
airport. It is common practice for these piofessionals to maintain or have access to detailed
historical data pertaining to the frequency of various wind directions in the locale of their
concern.
3-28
-------
4.0 FIRE HAZARDS OF CHEMICAL SUBSTANCES
4.1 INTRODUCTION
When most of us think of an unwanted fire, we typically picture a burning building, a
burning transport vehicle of some kind, or a burning forest with flames and smoke rising into
the sky. These are clearly the most common types of fires and typically involve ordinary
combustible materials such as paper, wood, cloth, plastics, and rubber Fire departments
across the nation face such fires on a daily basis and are well-equipped and trained to deal
with them. Hazardous materials, however, may pose additional types of fire hazards with
unusual characteristics In the following, we first discuss measures of flammability potential,
continue with a discussion of how the effects of fires may be evaluated, and finally, describe
a number of "special" types of fires associated with hazardous materials
4.2 MEASURES OF FLAMMABILITY POTENTIAL
It hardly needs saying, but most of us realize that some matenals are much more easily
ignited than others Some require only a spark, such as the propane or LP-gas fuel in a gas
barbecue, while others, such as a piece of granite, will not ignite even if placed under a
welding torch The most common measures of flammability potential for matenals which are
flammable or combustible are 1) flash points, 2) lower flammable or lower explosive limits;
3) upper flammable or upper explosive limits, and 4) autoigmtion temperatures These data
are readily available in various handbooks and hazardous material data bases when known,
and are commonly listed in chemical company material safety data sheets (MSDS) Fire
-------
safety and combustion experts may also consider ignition energy requirements, fire points,
flame spread rates, and heat and smoke generation rates of materials in evaluating then*
flammability characteristics, but knowledge of these latter attributes is not truly needed for
the purposes of this document and sources of appropriate data are not readily available to the
general public for a large number of subtances.
FlashPoints
The flash point of a combustible substance, in simple terms, is the lowest temperature
of a material at which the vapors over its liquid or solid surface will ignite and burn when
exposed to a specified ignition source without necessarily causing self-sustaining combustion
of the liquid or solid Flash points vary from temperatures far below zero degrees Fahrenheit
for flammable gases (such as natural gas, LP-gas, propane or butane), and volatile flammable
liquids (such as gasoline), to hundreds of degrees above zero for heavy fuel oils. (Note: The
temperature at which the vapors over a liquid or solid will ignite and continue to bum due to
self-sustaining combustion of a liquid or solid is called its fire point. These temperatures are
available in the professional literature for only a relatively few materials.)
Materials with low flash points relative to temperatures in the ambient (i e., natural)
environment are usually ignited easily by a spark (be it from metal scraping metal or stone or
from static electricity) or by a flame from any source. Most frequently, they are substances
that are normally gaseous at ambient temperatures or liquids that readily evaporate or boil
upon release. These vapors or gases can sometimes be earned by the wind to a source of
ignition somewhat distant from the discharge site of the material and flashback to the spill
source causing one or more of the fire hazards described later
Substances with flash point temperatures close to ambient temperatures are also easily
ignited by sparks or flames. The main difference between such materials and those described
in the previous paragraph is that the ignition source must be closer to the fuel in order for
ignition to take place. This follows from the observation that such materials are generally
liquids of lower volatility than materials with substantially lower flash points.
The higher the flash point temperature is above ambient temperature, the more difficult
it becomes to ignite a substance Under normal circumstances, a fuel with a high flash point
cannot be ignited by a spark or even a nearby flame unless: 1) the fuel is a liquid sprayed
into the air as a fine mist, 2) the fuel is a finely divided solid, 3) a portion of the fuel is
heated to its flash point by a nearby source of heat and then exposed to an ignition source; or
4) the fuel is heated to a temperature at or above its flash point prior to release and
encounters an ignition source before cooling
4-2
-------
The flash point temperatures of combustible materials are determined using testing
methods and equipment standardized by various organizations, with the American Society of
Testing and Materials (ASTM) being the primary standard-setting body in the United States.
There are two main classes of testing methods which respectively provide "open cup" or
"closed cup" flash points, and each class represents more than one specific testing method
Because of differences in equipment design and testing procedure, the numerical value of
closed-cup flash points is typically some 5-10° Fahrenheit lower than that of the open-cup
flash point for the same substance, but the difference may be greater or less in individual
cases Due to other factors, most importantly the purity of the sample tested, it is not
surprising to find a number of different closed cup or open cup flash points for any given
substance, all of which differ to some extent. It is well, therefore, to consider flash point
values reported in the literature as approximate rather than exact values
Flammability and Explosivity Limits
It is rather well known that combustion cannot take place in the absence of a certain
minimum amount of oxygen, be it available in air mixed with gases or vapors evolved from a
combustible substance or from an internal component of the fuel Conversely, there must be
sufficient fuel vapors or gases available in a fuel-air mixture to support and sustain
combustion. Thus, there are both lower and upper limits associated with fuel concentrations
in air that will ignite and permit flames to spread away from the source of ignition (i e,
permit flames to propagate) Fuel concentrations below the lower limit will contain
insufficient fuel to ignite and propagate flame and are commonly referred to as being too
lean to burn Those above the upper limit are considered too rich to ignite, that is, they
contain too much fuel and/or too little oxygen, as in the case of a "flooded" automobile
engine
The minimum concentration of a vapor or gas in air that will ignite and propagate
flame is known as its lower flammable limit (LFL) concentration or its lower explosive limit
(LEL) concentration and is usually expressed as a percentage by volume of fuel vapors in air.
The words flammable and explosive are used interchangeably such that LFL values typically
equal LEL values in the literature The reasoning behind this is that the concentration of a
fuel that will burn in air can also be expected to explode under the appropriate conditions
This supposition is only approximately true for some fuels (where precise LEL values might
be slightly higher than LFL values), but it has become widely accepted over decades of use
Similar to the above, the maximum concentration of a gas or vapor in air that will
ignite and propagate flame is known as the upper flammable limit (UFL) or upper explosive
limit (UEL) of the fuel. Again, the words flammable and explosive are often used in an
interchangeable fashion.
4-3
-------
LFL or LEL values are related to flash points of combustible substances in that the
flash point is theoretically the temperature at atmospheric pressure to which a substance must
be raised to produce a vapor or gas concentration over its surface equivalent to its LFL or
LEL concentration The relationship is not always observed in practice, however, because
flash point measurement equipment and procedures, as discussed above, do not always
produce precise values.
Flammable and explosive limits found in the literature are usually measurements made
at normal atmospheric temperatures and pressures unless indicated otherwise. Be advised
that there can be considerable variation in these limits at pressures or temperatures above or
below normal. The general effect of an increase in temperature or pressure is to reduce the
lower limit and increase the upper limit Decreases in temperature or pressure have the
opposite effect.
As a final note, it is also important to appreciate that certain solids, when dispersed in
air as fine powders, may also be capable of burning or exploding upon encountering a
suitable source of ignition. Some examples include coal dust produced in mining operations,
grain dust produced in silos during storage or transfer operations, and flour produced in
milling operations. Flammable or explosive limits for solid materials are usually expressed
in units of weight of solid present in a specified volume of air.
Autoignition Temperatures
The ignition or autoignition temperature (ATT) of a substance, whether solid, liquid,
or gaseous, is the minimum temperature necessary to initiate or cause self-sustaining
combustion in the absence of a flame or spark. Even more so than flash points or flammable
limits, these temperatures should be viewed as approximations due to the many factors that
can affect testing results. Indeed, it must be noted that most values currently found in the
literature were determined by testing methods that are now considered obsolete Newer
testing methods adopted by the ASTM frequently demonstrate substantially lower tempera-
tures for the onset of combustion than older methods.
Table 4.1 provides examples of various hazardous materials and their associated
flammability data Those at or near the top of the list are extremely flammable and volatile
and more likely to produce large quantities of flammable vapors or gases upon release,
vapors or gases that may travel a considerable distance from the spill site and still be within
flammable or explosive limit concentrations in an*. Those at or near the bottom of the list are
difficult to ignite without prior preheating and tend to have much lower vapor pressures (i.e.,
are generally of low volatility).
4-4
-------
TABLE 4.1
EXAMPLE FLAMMABILITY CHARACTERISTICS
Substance
Propane
Gasoline
Acetone
Isopropyl alcohol
Turpentine
Fuel oil no. 2
Motor oil
Peanut oil
Closed-cup
Flash Point (°F)
Very low
-45 to -36
-4
53
95
126-204
275-600
540
LFL (%)
21
1.4-1.5
25
20
0.8
*
*
*
UFL (%)
9.5
7.4-7.6
13
12.7 at
200°F
*
*
*
*
AIT(0F)
842
536-853
869
750
488
494
325-625
833
*Note: Flash points are often not recorded for substances that are gases
at ambient temperatures because of the very low temperatures
required to determine them. Similarly, flammable limits are not
always available for substances with high flash points due to
the high temperatures needed for ignition Substances that are
complex mixtures of a number of materials, (e g., fuel oils) may
have a range of flash points.
Sources: Fire Protection Guide on Hazardous Materials, 8th ed.,
National Fire Protection Association, Quincy, MA, 1984.
CHRIS Hazardous Chemical Data, U S Coast Guard, U.S.
Department of Transportation, Washington, D C, 1978.
4-5
-------
43 MEASURES OF FLAMMABILITY EFFECTS
Direct contact with a flame of any sort is obviously not a good idea for any prolonged
period of time since the extreme heat may ignite combustible materials or severely burn and
destroy living tissue What may not be fully realized is that fires can also cause damage or
injury from a distance via transmission of thermal radiation, not unlike the manner in which
the sun warms the earth. Such radiation, which is completely different from nuclear
radiation, will be strongest at the surface of a flame and will become rapidly weaker as one
moves away in any direction. Consequently, during a major hazardous material release
involving fire, property damage and human injuries may occur not only in burning areas, but
also in a zone surrounding the fire.
Thermal radiation levels (also referred to as thermal radiation fluxes) are measured and
expressed in units of power per unit area of the item receiving the energy. However, since
the damage or injury sustained by a receiving object is a function of the duration of exposure
as well as the level, thermal radiation dosages are also of concern These dosages are
determined by combining radiation levels with exposure times and are expressed in units of
energy per unit time per unit area of receiving surface Table 4 2 lists some of the known
effects of thermal radiation on bare skin as a function of exposure level and time.
4.4 TYPES OF FIRES
There are essentially six types of fires associated with hazardous material discharges,
with the type of fire a function not only of the characteristics and properties of the spilled
substance but the circumstances surrounding its release and/or ignition The six types are'
• Flame jets
• Fireballs resulting from Boiling Liquid Expanding Vapor Explosions
(BLEVEs)
• Vapor or dust cloud fires
• Liquid pool fires
• Fires involving flammable solids (as defined by the U S Department of
Transportation), and
• Fires involving ordinary combustibles
4-6
-------
TABLE 4.2
THERMAL RADIATION BURN INJURY CRITERIA
Radiation Intensity
kW/m1
1
2
3
4
5
6
8
10
12
Btu/hr-ft2
300
600
1000
1300
1600
1900
2500
3200
3800
Time for Severe
Pain (sec)
115
45
27
18
13
11
7
5
4
Time for 2nd
Degree Burn (sec)
663
187
92
57
40
30
20
14
11
Data sources:
Buettner, K, "Effects of Extreme Heat and Cold on Human Skin, n
Surface Temperature, Pain and Heat Conductivity in Experiments with
Radiant Heat," J. Ap Phys., Vol. 3, p. 703,1951.
Mehta, A K., et al, "Measurement of Flammabihty and Burn Potential
of Fabrics," Summary report to the NSF under Grant #GI-31881, Fuels
Research Laboratory, Mass. Inst of Tech, Cambridge, Mass., 1973.
4-7
-------
Flame Jets
Transportation or storage tanks or pipelines containing gases under pressure (ie.,
compressed gases) or normally gaseous substances that have been pressurized to the point
they become liquids (i e., compressed liquefied gases) may discharge gases at a high speed if
somehow punctured or broken during an accident The gas discharging or venting from the
hole will form a gas jet that "blows" into the atmosphere in the direction the hole is facing,
all the while entraining and mixing with air. If the gas is flammable and encounters an
ignition source, a flame jet of considerable length may form (possibly hundreds of feet in
length) from a hole less than a foot in diameter Such jets pose a thermal radiation hazard to
nearby people and property, and are particularly hazardous if they impinge upon the exterior
of a nearby intact tank containing a flammable, volatile, and/or self-reactive hazardous
material. Such events sometimes occur during multi-car train derailments or in incidents at
crowded chemical plants or oil/gas processing or storage facilities. In these cases, the heat of
the flame increases pressure in the intact tank while simultaneously weakening its outer wall.
This may eventually cause the tank to rupture violently or explode in an event referred to as a
BLEVE (see below), particularly if the flame impinges on the wall in the vapor space of the
container where there is no adjacent liquid to draw heat away from the wall surface. If the
contents of the intact tank are flammable, a large rising fireball may result If the contents
are nonflammable but toxic, a large amount of toxic vapors or gases may be suddenly
released to the atmosphere
Fireballs Resulting from BLEVES
Boiling Liquid Expanding Vapor Explosions (BLEVEs) are among the most feared
events when sealed tanks of liquid or gaseous hazardous materials are exposed to fire.
Although they are called explosions, they are not associated with strong blast waves in many
cases. Rather, they involve the violent rupture of a container of flammable material and the
rapid vaporization of the material If the substance is flammable, a large rising fireball may
form, the size of which will vary with the amount of hazardous material present, and which
may be as much as 1,000 feet in diameter when involving a railroad tank car containing a
flammable liquefied compressed gas like liquid propane or LPG Although the fireball is
generally of short duration, the intense thermal radiation generated can cause severe and
possibly fatal burns to exposed people over relatively considerable distances m a matter of
seconds. In addition, if the tank is relatively long and cylindrical in shape, part of the tank
may literally "rocket" into the air, all the while spewing forth burning gases and liquids.
Pieces of such tanks have been known to travel up to 5,000 feet in BLEVEs involving
railroad tank cars. Fires and various impact damages have occurred at the landing points of
larger pieces. (Note- Be advised that theie is potential for the tank to rocket upon rupturing
violently or exploding regardless of whether its contents are flammable or nonflammable)
4-8
-------
The phenomena leading to a BLEVE can occur with most hquids excessively heated in
a closed or inadequately vented container, whether they are flammable or not, or are pure
materials or mixtures, unless other circumstantial factors are considered. Two important
factors are the duration of the external exposure fire and the flow capacity of any pressure
relief valve if one is present If the exposure fire is not of sufficiently long duration, or if the
relief valve can vent vapor as fast as it is generated, a BLEVE will not occur An additional
factor is the availability of external cooling via fixed water spray systems, fire monitors, hose
streams, etc. These can contribute to the prevention of a BLEVE either by suppressing the
external fire or by cooling the heated vessel Finally, note that the possibility of a BLEVE
increases with the volatility of the hazardous material Substances with higher vapor
pressure at any given temperature are more at risk than those with lower vapor pressures
Vapor or Dust Cloud Fires
Vapors evolved from a pool of volatile liquid or gases venting from a punctured or
otherwise damaged container, if not ignited immediately, will form a plume or cloud of gas
or vapor that moves in the downwind direction. If this cloud or plume contacts an ignition
source at a point at which its concentration is within the range of its upper and lower
flammable limits, a wall of flame may flash back towards the source of the gas or vapor,
engulfing anything and everything in its path Similarly, fires may flash through airborne
clouds of finely divided combustible dusts whether or not they are formally classified as
hazardous materials. People or property caught within the cloud as the flame passes may be
severely injured or damaged if not protected
Liquid Pool Fires
A liquid pool fire is defined as a fire involving a quantity of liquid fuel such as gasoline
spilled on the surface of the land or water. As in pnor cases, primary hazards to people and
property include exposure to thermal radiation and/or toxic or corrosive products of
combustion. An added complication is that the liquid fuel, depending on terrain, may flow
downslope from the accident site and into sewers, drains, surface waters, and other
catchments. There have been cases where such fires have ignited other combustible
materials in the area or have caused BLEVEs of containers subjected to the flames On
occasion, pools of burning liquids floating on water have entered water intakes of industrial
facilities and caused internal fires or explosions. Burning fuels entenng sewers and drains
not completely full of fluid have caused underground fires and/or threatened industrial or
municipal treatment facilities at the receiving end of the sewer or drain.
4-9
-------
Flammable Solid Fires
A flammable solid is defined by the U S Department of Transportation as any solid
material, other than one classed as an explosive, which under conditions normally incident to
transportation is liable to cause fires through friction, retained heat from manufacturing or
processing, or which can be ignited readily and when ignited burns so vigorously and
persistently as to create a senous transportation hazard. Included in this class are sponta-
neously combustible and water-reactive materials
As the above definitions suggests, the term flammable solid encompasses materials
with a wide range of hazardous properties.
• Some of these solids are considered hazardous because they can be ignited by
friction, much like the head of a match
• Some are organic materials such as charcoal, powdered coal, wet paper, and even
fish scrap or fish meal which may at times internally generate heat to the point of
self-ignition when improperly stored or transported.
• Some are metals in the form of powders or other small pieces which can
self-ignite in prolonged contact with moisture, burn at very high temperatures,
and/or be difficult if not impossible to extinguish without special techniques or
materials, with aluminum and magnesium being good examples
• Some of these materials (i.e., pyrophonc substances) may ignite if exposed to air
or burn vigorously in the fashion of highway flares. Phosphorus has both of these
properties and also generates large quantities of toxic and irritating smoke.
• Some have several of these properties.
Fires Involving Ordinary Combustibles
Some hazardous materials, including some of the flammable solids described above,
burn with no special hazards beyond those associated with paper, wood, and other common
materials of everyday life Wet paper waste, for example, is only considered hazardous
because it may ignite spontaneously (i e, self-heat and self-ignite) Once burning, it poses
no special or unusual threat This is not meant to imply that such a fire would not be
significant or important to consider in planning for emergencies, only that the nature of the
threat is one encountered frequently by fire service personnel and well known to them.
4-10
-------
4.5 PRODUCTS OF COMBUSTION
Besides evolving heat and thermal radiation, fires involving certain hazardous materials
may generate smoke and gases that are more toxic than those evolved from ordinary
substances In most cases, the heat of a fire will cause these products of combustion to rise
into the sky where they will become diluted with air below harmful levels before
reapproaching the ground surface. On occasion, however, their toxicity level may be so high
as to necessitate public evacuations until the fire has been extinguished Indeed, a 1986
incident in Ohio involving the burning of phosphorus in a railroad car required the
evacuation of at least 40,000 people due to the toxic and irritating smoke generated This
was the largest evacuation associated with a train wreck in the history of the United States
Material safety data sheets (MSDS) and other data bases and handbooks describing
individual substances will typically provide a general indication of expected products of
combustion or thermal decomposition. The term "general" is used because far more often
than not the discussion will be rather imprecise and unlikely to highlight more than a few
rather common products of combustion or decomposition.
In the case of organic materials comprised solely of carbon, hydrogen, and oxygen,
products of combustion virtually always include carbon dioxide and highly toxic carbon
monoxide together with water vapor and some amounts of unburned vapors of the hazardous
material Substances of low molecular weight (i e., simple hydrocarbons and alcohols), may
indeed only generate these products of combustion when burning freely in the natural
environment More complex and heavier substances, however, may generate a complicated
mixture of substances, some of which may be extremely toxic A general rule of thumb to
follow is that most strictly organic materials usually pose no more hazard when burning
(although the hazard may indeed be very significant) than a burning wooden home or other
building The key exception involves fires involving organic materials of high toxicity in the
unburned state, with pesticides being primary examples. Fires involving such materials may
be particularly hazardous not only due to toxic combustion products but due to the potential
dispersion of unburned pesticides.
One can obtain a general idea of unusual products of combustion or decomposition by
looking at the chemical formula for any particular hazardous material of concern, this being
an item almost always given in MSDS and other safety related publications for pure
materials. Some of the more common symbols used for various individual components (i.e.,
elements) of chemical molecules include:
4-11
-------
Element
Bromine
Carbon
Chlorine
Fluorine
Hydrogen
Lead
Mercury
Nitrogen
Oxygen
Phosphorus
Sulfur
Chemical Symbol
Br
C
Cl
F
H
Pb
Hg
N
O
P
S
Hazardous materials containing bromine, chlorine, or fluorine, if subject to combustion
or decomposition in a fire environment, may generate irritating and corrosive substances
such as hydrogen bromide or hydrobromic acid, hydrogen chloride or hydrochloric acid, or
hydrogen fluoride or hydrofluoric acid, and possibly gaseous bromine, chlorine or fluorine
themselves. The extremely toxic substance known as phosgene may be formed in some cases
when chlorine is present, particularly in combination with oxygen in the chemical molecule,
so it is important to check for this possibility in MSDS and other information sources
Both lead and mercury are well-known toxic metals that can be found as components
of numerous chemical substances. Smoke or fumes from fires involving these toxic heavy
metals and others (such as arsenic), must always be of concern.
Although pure nitrogen gas is non-toxic and a major component of air, chemical
molecules containing nitrogen atoms may evolve toxic nitrogen oxides under fire conditions.
The combination of carbon with nitrogen in a -CN group within a chemical molecule
suggests that highly toxic cyanides may be generated in fires
Dry phosphorus may ignite upon contact with air and generate thick white smoke
containing phosphoric acid and phosphorus pentoxide. As noted earlier, this smoke is both
highly irritating and highly toxic.
4-12
-------
5.0 EXPLOSION HAZARDS OF CHEMICAL SUBSTANCES
5.1 DEFINITIONS
The dictionary contains two definitions of the word explode relevant to hazardous
materials, these being
• To burn suddenly so that there is violent expansion of hot gases with great
disruptive force and a loud noise (in what is called a thermal explosion).
• To burst violently as a result of pressure from within (in what is called a
non-thermal explosion).
The first definition clearly involves ignition and release of thermal energy from an
explosive material or mixture while the second does not In the following, we first discuss
the conditions and factors that define the potential for both thermal and non-thermal
explosions, follow with a discussion of how the effects of explosions can be measured, and
then discuss the various types of explosions which meet the above criteria and which may be
encountered in accidents involving hazardous materials
-------
5.2 FACTORS THAT INFLUENCE EXPLOSION POTENTIAL
Thermal Explosions
The definitions of lower and upper flammability limits presented earlier explained that
these terms are used interchangeably with the terms lower and upper explosive limits in air.
The reason for this is that a flammable mixture of a gaseous fuel in air, i e, a mixture within
the range of lower and upper flammable limit concentrations, may explode if ignited under
appropriate conditions. Similarly, a cloud of combustible dust may explode if airborne
concentrations are within these limits and the cloud is confined
The set of conditions under which explosions of gases or vapors are most common
involves ignition within the confined space of a building, sewer pipe, tunnel, partially empty
liquid storage tank (on land or on a marine vessel), or other container. Dust explosions have
frequently occurred in grain handling facilities and storage silos as well as other locations
where fine combustible dusts are handled or generated.
It follows from the above that virtually all substances that are handled under conditions
in which fuel-air mixtures are within explosive or flammable limits and fill a significant
fraction of an enclosed space have a high probability of exploding rather than simply burning
upon ignition. However, it must also be leahzed that gaseous mixtures may also explode at
times when only partially confined or even if completely unconfined in an open environment
These latter explosions, referred to as unconfined vapor cloud explosions, often have far less
power than explosions in confinement, and it has been observed that some substances have a
far greater probability of exploding when unconfined than others Nevertheless, past events
have proven that unconfined explosions can occasionally cause devastating damage and
widespread injuries, especially when the weight of airborne gas or vapor exceeds 1000 Ibs
Below this weight, unconfined vapor cloud explosions are quite rare and typically involve a
relatively few specific materials.
There also are many solids and liquids which may explode or detonate if ignited,
shocked, or subjected to heat or friction, depending on their individual properties and
characteristics. Some of the best known examples are TNT, dynamite, gunpowder, and
nitroglycerine which may be referred to at times as condensed-phase explosives or high
explosives Determination of whether any particular liquid or solid may be explosive, and
the conditions under which it may explode, requires investigation on a case-by-case basis,
since there is no specific property or characteristic that sets explosives apart from other
materials Fortunately, manufacturers of these materials and hazardous matenal data bases
and guidebooks will usually highlight the explosive properties of such materials
The power or strength of a thermal explosion, however one wishes to express it, is a
function of three primary factors:
5-2
-------
• The amount of fuel present that is capable of exploding
• The amount of energy available in this portion of the fuel
• The fraction of the available energy (known as the yield) expected to be
released in the explosion process.
In simpler terms, it is understandable that two sticks of dynamite produce a larger blast
than one stick, that fuel-air mixtures above or below explosive limit concentrations in air
may not give additional strength to an explosion, and that some substances contain more
energy per unit weight than others
Non-thermal Explosions
The most simple type of non-thermal explosion to understand is that due to
overpressunzation of a sealed or inadequately vented container of some sort. Much as a
balloon will burst if too much air is blown in, the walls of a sealed tank or other container
may rupture violently if too much gas or liquid is forced in, if an internal chemical reaction
produces excessive gases or vapors, or if a reaction or other source of heat increases the
internal vapor pressure of the contents to the point that the walls are "stretched" beyond their
breaking point. Since ignition and fire are not involved in the actual explosion process, these
events are considered non-thermal explosions, although the contents of the container may
ignite subsequent to its release if a suitable ignition source is present and the substance is
flammable or combustible.
The strength of a non-thermal tank overpressunzation explosion is a function of the
pressure at which the walls of the container burst and the nature of the walls (i e, whether
they are brittle and will break suddenly with a "snap" or are ductile and more likely to stretch
and then split or tear along some line on the surface). If the tank contains gas under pressure,
the volume of the gas in the tank will also be important.
A final note is that non-thermal explosions involving compressed gases or vapors are
far more likely to cause damage to distant objects than those involving liquids This follows
from the definition of shock and blast waves presented below and the relatively incompress-
ible nature of liquids.
5.3 MEASURES OF EXPLOSION EFFECTS
When a firecracker or a stick of dynamite explodes, the violence and speed of the
reactions taking place produce what is either referred to as a shock wave or a blast wave.
Technically speaking, there is a difference between these two terms, but we will treat them
rather interchangeably here. Either type of wave can be thought of as a thin shell of highly
5-3
-------
compressed air and/or hot gases that rapidly expands in all directions from the point at which
the explosion is initiated Such waves can move at velocities exceeding the speed of sound
in air, and, therefore, are capable of producing sonic "booms," much like those associated
with supersonic aircraft. This is why significant explosions produce a loud "bang "
The damage caused by a shock or blast wave striking an object or a person is a
complex function of many factors, and it is well beyond the scope of this document to
attempt to describe all the complex interactions involved. Instead, we will simply refer to the
wave as a rapidly expanding shell of compressed gases The strength of the wave can then be
measured in units of pressure (psi, e g), and the effects of peak overpressures within the
wave (i.e., the maximum pressure in the wave in excess of normal atmospheric pressure) can
be related to the level of property damage or personal injury likely to result
Table 5.1 presents a list of peak overpressures and their expected effects on people and
property. It is important to note that peak overpressures in a shock or blast wave are highest
near the source of the explosion and decrease very rapidly with distance from the explosion
site. Additionally, it must be noted that the location of the blast relative to nearby "reflecting
surfaces" will influence the extent of damage incurred. For example, picture an explosion
that takes place well above the surface of the ground. In this type of elevated or "free-air"
event, the spherical shock wave has the opportunity to travel and dissipate in all directions
simultaneously. Conversely, if the same explosion were to take place directly on the ground
surface, the major portion of the energy released would only dissipate upwards and outwards
The ground surface would reflect most energy directed downward, and the net result would
be a blast or shock wave with approximately twice the strength expanding from a
hemi-sphencal shaped volume of space situated on the ground Hazard analysis procedures
discussed in Chapter 12 and Appendix B of this guide and incorporated into the ARCHIE
Computer Program therefore consider the location of an explosion relative to the ground
surface. Not considered, however, are potential reflections from building walls and other
surfaces that may cause actual damage patterns to be somewhat more erratic than those
predicted by generalized hazard assessment methodologies for explosion events
Beside personal injuries and property damage caused by direct exposure to peak
overpressures, the blast or shock wave also has the potential to cause indirect impacts. These
secondary effects of explosions include:
• Fatalities or injuries due to missiles, fragments, and environmental debns set in
motion by the explosion or by the heat generated.
• Fatalities or injuries due to forcible movement of exposed people and their
subsequent impact with ground surfaces, walls, or other stationary objects
5-4
-------
TABLE 5.1
EXPLOSION OVERPRESSURE DAMAGE ESTIMATES
Overpressure*
(psig)
003
004
010
015
030
040
050-10
07
10
10-20
10-80
13
20
20-30
23
24-122
25
30
30-40
40
50
50-70
70
70-80
90
100
14 5-29 0
Expected Damage
Occasional breaking of large windows already under stress
Loud noise (143 dB), sonic boom glass failure
Breakage of small windows under strain
Typical pressure for glass failure
Some damage to house ceilings, 10% window glass breakage
Limited minor structural damage
Windows usually shattered, some window frame damage
Minor damage to house structures
Partial demolition of houses, made unmihabitable
Corrugated metal panels fail and buckle Housing wood panels blown in
Range for slight to serious injuries due to skin lacerations from flying glass and other missiles
Steel frame of clad building slightly distorted
Partial collapse of walls and roofs of houses
Non-reinforced concrete or cinder block walls shattered
Lower limit of serious structural damage
Range for 1-90% eardrum rupture among exposed populations
50% destruction of home brickwork
Steel frame building distorted and pulled away from foundation
Frameless steel panel building ruined
Cladding of light industrial buildings ruptured
Wooded utility poles snapped
Nearly complete destruction of houses
Loaded tram cars overturned
8-12 in thick non-reinforced bnck fail by shearing of flexure
Loaded tram box cars demolished
Probable total building destruction
Range for 1-99% fatalities among exposed
populations due to direct blast effects
*These are the peak pressures formed in excess of normal atmospheric pressure by blast and shock waves
Source Lees J'P, Loss Prevention m the Process Industries. Vol 1, Butterworths, London and Boston, 1980
5-5
-------
The most common injuries due to missiles and the like are attributable to violent glass
breakage and impact of airborne shards of glass with people. Fragments may include
portions of any container that explodes and pieces of structures or equipment that are torn
loose by the explosion and become airborne. Environmental debris essentially covers all else
that may be forced out of place The entire category can also be considered to encompass
situations in which people are buried in the rubble of collapsed buildings or other structures.
It is very important to realize that a tank that BLEVEs or otherwise ruptures violently
may break up into various fragments, one or more of which may be projected for
considerable distances Portions of cylindrical tanks have been known to literally "rocket"
into the air while spewing forth burning liquids and have caused fires and impact damages
upon falling back to the ground.
Where railroad tankcars or highway tank vehicles are at nsk, hazardous material
response guides have typically suggested that a radius of one-half mile be evacuated to
prevent injuries from both fragment and thermal radiation hazards Recent incidents have
indicated, however, that individual fragments may occasionally travel as far as 4000-5000
feet from a tankcar BLEVE, and it is therefore prudent to evacuate to a radius of one mile in
such cases, if this is practical Since railroad tankcars carry 2-4 times as much cargo as
typical highway tank vehicles, the one-half mile radius improbably sufficient for major truck
accidents, but this is not absolutely certain for all cases
The evacuation distances required for smaller or larger tanks than typically
3,000-12,000 gallon highway vehicles or 20,000-30,000 gallon capacity railroad tankcars
will vary somewhat with the quantity of hazardous material present, but not as much as one
might think. At the lower end of the scale, one major authority suggests a 1500 ft evacuation
radius for situations in which an ordinary gas cylinder is involved in fire. Limited data for
explosions or BLEVEs involving major stationary storage tanks do not indicate fragment
hazards beyond one mile in the majority of known cases.
Where a tank or container ruptures violently due to internal overpressunzation,
fragment hazards are to some degree a function of whether the wall materials are battle or
ductile. Brittle materials (such as glass) may shatter into many smaller pieces Tanks or
containers made of ductile materials (such as most metals at or above relatively normal
temperatures) are more likely to split or tear into a few large pieces
Fatalities or injuries due to forcible movement of exposed people and then- subsequent
impact with objects quite literally involves situations in which the shock or blast wave pushes
or picks up and throws bodies against obstacles
5-6
-------
5.4 TYPES OF EXPLOSIONS
Many of the basic types of explosions have already been described, but there are
benefits in listing them again and providing more formal definitions of terms.
Container or Tank Overpressurization Explosions
As noted earlier, these events are a result of excessive pressure within a sealed tank or
other container and are deemed to be non-thermal explosions They occur when excessive
pressure causes the walls of a tank or container to rupture violently, much as a balloon
"pops" when too much air is blown in.
Dust Explosions
A cloud of combustible dust that is airborne and has concentrations within its upper
and lower explosive limits may explode when ignited Explosions usually occur when the
dust fills most of an enclosed space of some kind.
An earlier discussion of fire hazards described how non-exploding clouds of dust in air
may simply burn in a dust cloud fire that can also be referred to as a deflagration It is
important to realize that there is no fine line between a deflagration and an explosion, since
deflagrations are also capable of producing shock waves with measurable peak overpres-
sures. It is usually when these overpressures become significant to the point of causing
damage or injury that the event is called an explosion It is when the shock or blast wave
moves at a velocity greater than the speed of sound under the conditions present, thus being
capable of causing maximum damage, that the event may be called a detonation
Gas or Vapor Explosions
As in the case of airborne dusts, a gas or vapor within flammable or explosive limit
concentrations may cause a deflagration, explosion, or detonation upon ignition. These
events can occur when the fuel-air mixture is confined, partially confined, or completely
unconfined, but confinement of the mixture most definitely increases the probability of
significant personal injury or property damage Note that the gas or vapor may be directly
released to the vulnerable environment or may evolve from evaporating or boiling liquids
that have entered the area
5-7
-------
Condensed-Phase Explosions or Detonations
As noted above, when the substance that explodes or detonates is a liquid or a solid, the
event is often called a condensed-phase explosion or detonation Those who use this term
may be prone to call events involving gases or vapors in air as diffuse-phase or gas-phase
explosions or detonations.
Boiling Liquid Expanding Vapor Explosions (BLEVEs)
BLEVEs were described in some detail in the prior section discussing fire hazards of
concern, where it was stated that they are not associated with strong shock or blast waves in
many cases. Obviously, this also means that shock or blast waves with sufficient power to
cause injury or damage may indeed occur at times.
Although some experts may disagree with the fine points of what is being said,
BLEVEs can also be described as a combination of other types of fires and explosions.
Indeed, bursting of a tank of liquid or compressed liquefied gas due to overheating is related
to tank or container overpressunzation explosions. Subsequent ignition of expanding gases,
which may result in a large fireball, can be thought of as resulting in one type of gas or vapor
cloud deflagration.
5-8
-------
6.0 TOXICITY HAZARDS OF CHEMICAL SUBSTANCES
6.1 INTRODUCTION
Although hazardous materials can pose both short-term and possibly long-term
lexicological threats to terrestrial and aquatic wildlife and plants, the immediate concern
during significant discharges is protection of human life and health Consequently, this
section addresses the toxicity and toxic hazards posed to the public by chemical substances
It must be noted, however, that much of what will be presented can also be applied to
understanding lexicological hazards to plants and animals
62 ROUTES OF ENTRY
Toxic materials, be they solids, liquids, or gases/vapors, can affect living creatures via
three primary routes of entry.
• Inhalation -- the process by which irnlanls or loxins enter ihe body via the
lungs as a resull of the respiratory process
Ingestion — the process of consuming conlaminaled food or water or
olherwise permitting oral inlake of imlanls or loxins
-------
• Direct contact with skin or eyes ~ the process by which hazardous
materials cause injury to bodily tissues via direct contact 01 cause poison-
ing via absorption through the skin or other external tissues Also included
in this category is the passage of toxic materials into the body via puncture
wounds or other breaks in the skin
Inhalation exposures may result from breathing gases vented from containers, vapors
generated from evaporating liquids (on land or in water), liquid aerosols generated dunng
venting of pressurized liquids, fumes generated from spilled acids, gases or fumes generated
by chemical reactions, dusts that become airborne due to an explosion or due to wind forces,
the products of combustion of a burning hazardous material, or a variety of other
mechanisms.
Ingestion (i e., oral) exposures may follow from poor hygiene practices after handling
of contaminated materials or from ingestion of contaminated food or water. Ingestion may
also occur following inhalation of insoluble particles that become trapped in mucous
membranes and swallowed after being cleared from the respiratory tract
Direct contact may result from exposures to hazardous gases, liquids or solids in the
environment, either on land, in the air, or in water Effects may be local and involve
irritation or burns of the skin or eyes or involve poisoning via absorption through external
bodily tissues.
The fact that a toxic chemical can cause harm by inhalation, mgestion, or irritation or
burning of the skin or eyes is probably well appreciated by most people Poisoning due to
absorption through external bodily tissues, however, is not as well known a hazard and
benefits from further explanation.
In simple terms, there are various specific gases, liquids, and even solid materials
which have the capability of passing thiough the skin or tissues of the eyes at various rates
upon contact Those that are highly toxic and which penetrate the body rapidly are the most
hazardous Those that penetrate slowly or which are of relatively low toxicity may require
long term contact with large parts of the body to cause significant effects. Although some
materials may give some warning that contact has occurred by causing some sort of burning
sensation, others may give little or no warning to the victim.
While on this topic, it is also worthwhile to consider the commonly accepted meaning
of phrases hke high toxicity and low toxicity When one speaks of a material that is of high
toxicity, it generally means that relatively small quantities may cause significant health
effects upon inhalation, ingestion, and/or direct contact Conversely, a low toxicity substance
generally requires larger amounts to be inhaled, ingested, or contacted for an equally
significant adverse health effect It is therefore well to always remember that a large quantity
6-2
-------
of a low toxicity material may present the same or greater toxic hazard to a community or
individual than a much smaller quantity of a highly toxic material. It is also necessary to
understand that the toxicity of a material is only one of several factors to be considered in
determining the toxic hazard posed by the material. These concepts are reiterated and
discussed in further detail in a later section.
6.3 TYPES OF TOXIC EFFECTS
Most toxic substances can be classified as irritants, asphyxiants, anesthetics and
narcotics, systemic poisons, sensitizers, carcinogens, mutagens, and/or teratogenic sub-
stances. Systemic poisons may be further disaggregated into the categones of hepatotoxic
agents, nephrotoxic agents, neurotoxic agents, agents which act on the blood or hematopoiet-
ic system, and agents which damage the lung.
Many of these terms may be unfamiliar because they are mostly used in the
medical/public health community and among lexicologists. Fortunately, they need not all be
memonzed because most hazardous material data bases and guides, material safety data
sheets, and manufacturers' product bulletins generally "translate" the effects of toxic
materials upon the body into more common language. There are, however, certain terms and
expressions that appear frequently and which can be helpful in understanding the most
common effects of toxic materials upon the body.
Irritants
Irritants are substances with the ability to cause inflammation or chemical burns of the
eyes, skin, nose, throat, lungs, and other tissues of the body in which they may come in
contact. Some substances such as strong acids (eg, sulfunc acid, oleum, chlorosulfonic
acid, hydrochloric acid, hydrofluoric acid, or nitric acid) may be irritating to the point of
being corrosive when concentrated, and may quickly cause second or third degree chemical
burns upon contact with the skin or eyes. If inhaled as a gas, vapor, fume, mist, or dust, they
may cause severe lung injury, and if ingested, can senously damage the mouth, throat,
stomach, and/or intestinal tract. Yet other irritants may have milder effects and may only
cause reddening of the skin or eyes after contact.
Some of the most common irritants are organic solvents or hydrocarbon fuels which
can dissolve natural oils ui the skin and cause dermatitis. After repeated or prolonged
contact, these will dry the skin to the point that it may become cracked, inflamed and
possibly infected. These same materials often cause irritation of the eyes and possibly loss
upon contact of the cornea! epithelium, a clear thin membrane that covers the surface of the
cornea. Although the effect is temporary, since the epithelium will usually regrow in a few
days, some data sources may refer to the effect as a "cornea! burn."
6-3
-------
Entry into the lungs of many kquid hydrocarbons and some organic liquids that are
irritants may cause chemical pneumonia or pneumonitis together with pulmonary edema
(filling of the lungs with fluid), hemorrhage, and tissue necrosis (i e., death of living tissue)
Since entry of liquids into the lungs usually involves aspiration when a victim who has
accidentally ingested the substance vomits, the first aid instructions for such substances
typically recommend against intentional inducement of vomiting. They also are likely to
mention that the effects of aspiration into the lungs may not appear for several hours or even
days after the exposure has taken place
Asphyxiants
Simple asphyxiants are typically non-toxic gases that may cause injury by inhalation
only if they are present hi air in such high concentrations that they displace and exclude the
oxygen needed to maintain consciousness and life A good example is nitrogen, a gas that
makes up about 78% of the air we breathe and which is perfectly harmless at this level as a
component of air. If additional nitrogen or another such simple asphyxiant were added to the
air to the point that the normal oxygen concentration of approximately 21 percent by volume
was significantly reduced, however, the situation could become life-threatening Tables 6 1
and 6.2 illustrate the effects of oxygen depletion on the body and the four stages of
asphyxiation.
Chemical asphyxiants are substances that in one way or another prevent the body from
using the oxygen it takes in and are often highly toxic substances. One classic example is
carbon monoxide which combines with and "ties up" the component of blood (hemoglobin)
that transports oxygen from our lungs to other organs If too much of the hemoglobin
becomes unavailable for carrying oxygen, a person may pass out and eventually die Other
examples are among the family of cyanides (i e, substances which have a -CN, carbon-nitro-
gen, combination in their molecule and which somewhere in their names have the word
"cyanide" or the letter combinations "cyan" or "nitrile"). These act by interfering with the
action of the enzymes necessary for Irving tissues to use available oxygen, thus resulting in a
condition referred to as cyanosis.
Anesthetics and Narcotics
Numerous hydrocarbon and organic compounds classified as hazardous materials,
including some alcohols, act on the body by depressing the central nervous system (CNS)
Early symptoms of exposure to these substances include dizziness, drowsiness, weakness,
fatigue, and incoordination. Severe exposures may lead to unconsciousness, paralysis of the
respiratory system, and possibly death
6-4
-------
TABLE 6.1
EFFECTS OF OXYGEN DEPLETION
Percent of Oxygen
In Air
20
17
12 to 15
10 to 12
6 to 8
6 or below
Symptoms
Normal
Respiration volume increases, muscular coordination diminishes,
attention and clear-thinking requires more effort
Shortness of breath, headache, dizziness, quickened pulse, efforts
fatigue quickly, muscular coordination for skilled movements lost
Nausea and vomiting, exertion impossible, paralysis of motion
Collapse and unconsciousness occurs.
Death in 6 to 8 minutes
Source Kimmerle, George, "Aspects and Methodology for the Evalution of Toxicological
Parameters During Fire Exposure," JFF'/Combustion Toxicology, Vol. 1, February,
1974
TABLE 6.2
FOUR STAGES OF ASPHYXIATION
1st Stage
2nd Stage
3rd Stage.
4th Stage:
21-14% oxygen by volume, increased pulse and breathing rate with
disturbed muscular coordination.
14-10% oxygen by volume, faulty judgment, rapid fatigue, and
insensitivity to pain
10-6% oxygen by volume,nausea and vomiting, collapse.and perma-
nent brain damage
Less than 6% by volume, convulsion, breathing stopped, and death
Source Cryogenics Safety Manual, Bntish Cryogenics Council, London, 1970
5-5
-------
Sensitizers
A few hazardous materials are sensitizers and cause sensitization. What this means is
that some people who are exposed to one of these materials may not be abnormally affected
the first time, but may expenence significant and possibly dangerous effects even in the
presence of very low levels of the contaminant if ever exposed again. In simple terms,
victims become extremely allergic to the material and possibly others of a similar nature
Other Types of Toxic Agents
• Hepatotoxic agents are materials that cause liver damage
• Nephrotoxic agents are materials that cause kidney damage
• Neurotoxic agents are substances that in one way or another impact the
nervous system and possibly cause neurological damage.
• Carcinogens are substances that may incite or produce cancer within some
part of the body.
• Mutagens can produce changes in the genetic material of cells.
• Teratogenic materials may have adverse effects on sperm, ova, and/or fetal
tissue.
Note: Besides the chemical asphyxiants described above, there are other substances that in
one way or another act on the blood or the hematopoietic system (i.e., bone marrow).
Inhalation of free silica or asbestos over a period of time can cause changes in lung tissue
with serious health consequences Yet other toxic substances also have unusual or unique
effects on human health.
6.4 ACUTE VS. CHRONIC HAZARDS
The majority of industries and many common daily activities of life utilize equipment,
processes, and materials that continuously or intermittently discharge toxic materials into the
occupational and/or natural environment Some workers may be exposed to such materials 8
hours per day, 5 days per week or so, over a large part of their careers. Similarly, the general
public may be exposed to various contaminants continuously or intermittently. Such
exposures are said to be of a chronic nature and usually but not always involve low
concentrations of contaminants in air, food, water, and/or soil.
6-6
-------
When a major accident or other rare event causes a significant spill or discharge of a
toxic material into the environment, the general public or nearby workers may be exposed to
relatively high levels of one or more toxic contaminants until such time as they escape or are
rescued from contaminated locations or the contaminant becomes diluted below hazardous
levels. These short-term, rare exposures (in the sense there will be long periods of time
between repeated exposures if they reoccur at all) are referred to as acute exposures Not all
acute exposures, of course, need involve high concentrations of toxic materials A small spill
or discharge may produce low levels of contamination yet still be of an acute nature
To be noted is that many chemicals will not persist for long periods of time in the
environment, or at least in those parts of the environment of concern, while others may
remain present for weeks, months, or even years. The former materials include substances
that may be digested by bacteria (i e, which are biodegradable), substances that will undergo
various reactions with materials in the environment that render them harmless, or those that
become so diluted in air or water that they no longer present a hazard Examples are simple
alcohols that may be digested by bacteria in soil or water much as humans drink and digest
alcoholic beverages, as well as volatile materials which evaporate and are swept away into
the vast ocean of air above us. Such materials are unlikely to pose long-term chronic hazards
in the event of a major spill or discharge in most cases Alternatively, toxic substances which
are relatively inert and which do not degrade, react, vaponze, or dissolve freely may pose
health hazards for extended periods of time within a localized environment and may require
additional planning to address long-term chronic exposure hazards to the public Examples
include heavy metals and various chlorinated hydrocarbons such as DDT, tnchloroethylene,
and PCBs
6.5 IMPORTANCE OF EXPOSURE LEVEL AND DURATION
In considering the effects of toxic exposures, it is necessary to understand that the
duration of an exposure can be as important as the level of exposure in determining the
outcome This follows from the observations that:
• The body has a capacity to cope with the intake of many contaminants at a
certain rate Below a certain threshold rate of intake or absorption which
can be counterbalanced by the body's ability to excrete or somehow convert
the contaminant to a harmless substance, toxic effects may be minimal or
non-existent For example, note that arsenic is commonly found in all
human bodies at low levels. It is only when the level exceeds the safe
threshold due to excessive intake that symptoms of toxicity become
apparent.
6-7
-------
• The rate at which a contaminant enters the body by inhalation is a function
of the concentration of the contaminant in the air being breathed, the rate of
breathing, the length of time the body remains within a volume of
contaminated air, and the specific properties of the contaminant. Higher
concentrations in air obviously lead to higher rates of intake or absorption
into bodily tissues.
• The potential for toxic effects via skin absorption is a function of the
amount of toxic material that contacts the body, the properties of the
material, and the length of time it is permitted to remain in contact
• Toxic effects via ingestion can also be a function of the amount or rate of
intake over a period of time. Small doses of certain poisons ingested hours
or days apart may not be harmful, but taking the total amount all at once
may be deadly. Other poisons may accumulate in the body such that small
doses taken over time may buildup to a fatal dose.
The reason that chronic exposure to low levels of toxic materials commonly found in
the environment does not often cause widespread health problems is that the rate of intake is
below the threshold at which health effects become apparent Conversely, major spills or
discharges of toxic materials may pose a significant threat to public health because the
resulting contaminant concentrations in the local area may be so high that only a moment or
two of exposure is sufficient to produce severe health problems due to an excessive body
burden of contamination. This is particularly true where large amounts of toxic gases or
vapors are released into the air. Relatively few members of the general public are ever
harmed by direct contact with toxic materials, since most individuals have the common sense
not to touch or go walking through spilled chemicals and will cleanse themselves promptly if
such contact is made. Similarly, few people are likely to drink potentially contaminated
water or eat contaminated food once warned of the possibility of contamination Most at risk
in such situations are emergency response personnel who enter contaminated areas without
adequate personnel protective clothing and respiratory devices in attempts to contain or
otherwise mitigate the impacts of the spill
6.6 TOXICITYVS. TOXIC HAZARD
The observations above naturally lead to a further discussion of the difference between
the toxicity of a substance and the toxic hazard it poses to the public This is an extremely
important concept because materials of high toxicity are often assumed to pose a severe toxic
hazard regardless of the other properties of the material and the circumstances surrounding
its spillage.
6-8
-------
Imagine a one ton discharge of two different materials. The first is an extremely toxic,
non-volatile solid material that has spilled in the middle of a street in a densely populated
metropolis. The material is so extremely toxic that only 10 pounds would be sufficient to kill
100,000 people by mgestion if somehow introduced into their food in equal portions. The
second discharge involves an overturned tank truck on the same street that has just released a
very common compressed liquefied gas that is considered to be of moderate toxicity. As it
boils and vaponzes upon release, the ton of liquid may become as much as 30,000 cubic feet
or more of pure gas If it were to mix uniformly with air and happened to be deadly in very
short-term exposures at a concentration of 5,000 ppm in air, the potentially lethal cloud
spreading over the city would conceivably have a total volume of 6 million cubic feet
On a strictly weight basis, the solid material may be many thousands of times more
toxic than the gas, but is unlikely to poison members of the public just a short distance away
because it lacks mobility Thus, the solid must be carefully handled and removed from the
scene, but actually poses a relatively low toxic hazard to the public Authorities may wish to
evacuate the immediate spill area and cover the solid with plastic sheeting to prevent any
dust from becoming airborne until its careful recovery, but the risk of fatalities among the
general public will be low in most cases.
The situation with the lower toxicity liquefied gas poses a greater toxic hazard because
the gas will quickly spread over downwind areas. The gas may prove rapidly fatal to people
near the spill site and cause toxic effects among many hundreds or thousands of others in the
downwind direction.
The moral of this story is that the toxic hazard posed by a material is not a sole
function of its toxicity. One must always consider the amount of material present or spilled,
the properties of the substance, and the opportunity it has to affect the population in its
vicinity.
6.7 RECOGNIZED EXPOSURE LIMITS FOR AIRBORNE CONTAMINANTS
It should be fairly clear by this point that discharges of gases and vapors into the
atmosphere generally pose greater toxic hazards to people than discharges of non-volatile
materials As is widely appreciated, one of the key tasks in planning for hazardous materials
emergencies involves preparations for identifying, notifying, evacuating, sheltering, or
otherwise protecting populations that may be exposed to such gases and vapors
Achievement of the above goal requires planning personnel to select the airborne
concentration in air that can be tolerated by exposed populations while toxic vapors or gases
remain in the immediate area, since it is this concentration that will determine the boundaries
of the hazard zone. This, in turn, requires knowledge of the source and nature of commonly
6-9
-------
available and accepted exposure limits for airborne contaminants as well as their various
advantages and disadvantages for the intended use Primary data sources to be considered
include:
ACGIH Threshold Limit Values (TLVs)
• OSHA Permissible Exposure Limits (PELs)
• AIHA Workplace Environmental Exposure Limits (WEELs)
• NIOSH Immediately Dangerous to Life or Health Levels (IDLHs)
• AIHA Emergency Response Planning Guidelines (ERPGs)
• NAS/NRC Emergency Exposuie Guidance Levels (EEGLs) and Short-term
Public Emergency Guidance Levels (SPEGLs)
ACGIH TLVs
The American Conference of Governmental Industrial Hygiemsts (ACGIH) formed a
committee in 1941 to review available data on toxic compounds and to establish exposure
limits for employees working in the presence of airborne toxic agents The committee
continues to this day to publish an annual list of several hundred compounds and
recommended exposure limits in a booklet titled Threshold Limit Values and Biological
Exposure Indices. Copies of the latest edition were available for $5 in late 1988 from the
ACGIH at 6500 Glenway Ave, Bldg D-7, Cincinnati, Ohio 45211 or (513) 661-7881.
The primary purpose of the exposure limits adopted by the ACGIH is to protect healthy
male workers in chronic exposure situations and the ACGIH specifically notes that "These
limits are not fine lines between safe and dangerous concentrations nor are they a relative
index oftoxicity, and should not be used by anyone untrained in the discipline of industrial
hygiene." Nevertheless, the information provides valuable guideposts for identifying expo-
sure limits that will usually be decidedly safe for short-term acute exposures
Exposure limits established and published by the ACGIH are of several different types
and include:
Threshold Limit Value - Time Weighted Average (TLV-TWA). The time
weighted average concentration for a normal 8-hour workday and a 40-hour
workweek, to which nearly all workers may be repeatedly exposed, day
after day, without adverse effect
6-10
-------
Threshold Limit Value-Short Term Exposure Limit (TLV-STEL) A
time-weighted average concentration to which workers should not be
exposed for longer than 15 minutes and which should not be repeated more
than four times per day, with at least 60 minutes between successive
exposures This limit supplements the TLV-TWA where there are recog-
nized acute effects from a substance whose toxic effects are primarily of a
chronic nature. STELs are recommended only where toxic effects have
been reported from high short-term exposures in either humans or animals
Threshold Limit Value-Ceiling (TLV-C): The concentration in air that
should not be exceeded during any part of the working exposure Ceiling
limits may supplement other limits or stand alone.
In addition to the above limits, the ACGIH occasionally enters the notation "skin" after
listed substances. This notation indicates the potential for absorption of the substance
through the skin, eyes, or other membranes and the possibility that such absorption may
contribute to the overall exposure. An excessive amount of absorption may invalidate any
TLV limit, a high potential for direct contact with the substance may suggest the need for
special protective measures.
For many of the materials with an assigned TLV-TWA, the ACGIH could not find
sufficient lexicological data to establish a TLV-STEL For these substances, it recommends
"Short-term exposures should exceed three times the TLV-TWA for no more than a total of 30
minutes during a work day and under no circumstances should they exceed five times the
TLV-TWA, provided that the TLV-TWA is not exceeded" for the 8-hour workday The
airborne concentrations denved from this recommendation are referred to as excursion
limits
OSHAPELs
The Occupational Safety and Health Administration (OSHA) within the U S Depart-
ment of Labor is responsible for the adoption and enforcement of standards for safe and
healthful working conditions for men and women employed m any business engaged in
commerce m the United States. When first established in the early 1970's, OSHA essentially
adopted the then current ACGIH TLV-TWAs and TLV-Cs as occupational exposure limits
and made them official federal standards Instead of calling the limits Threshold Limit
Values, however, it referred to them as Permissible Exposure Limits (PELs) As in the case
of TLVs, there are both time-weighted average (TWA) and ceiling (C) values for various
materials as well as occasional peak values for shorter time penods While the ACGIH
reviews and frequently revises its TLVs on an annual basis, OSHA did not similarly update
6-11
-------
its PELs except for a relatively small number of individual substances until early 1989 when
it lowered the PELs for 212 widely used chemicals, adopted new PELs for 164 substances
not previously regulated, and reaffirmed the PELs for 52 materials.
PELs are formally listed in Title 29 of the Code of Federal Regulations (CFR), Part
1910, Subpart Z, General Industry Standards for Toxic and Hazardous Substances An
inexpensive and valuable source of current PELs and much other information on chemical
hazards is the NIOSH Pocket Guide to Chemical Hazards published by the National
Institute for Occupational Safety and Health, a part of the U S Department of Health and
Human Services. When in stock, single copies may be available at no cost from NIOSH
Publications, 4676 Columbia Parkway, Cincinnati, Ohio 45226 (Telephone- 513-533-8287)
Copies are otherwise available at nominal cost as DHHS (NIOSH) Publication No 85-114
from the Superintendent of Documents, U.S Government Printing Office, Washington, D C
20402 or one of the many regional branches of the GPO. Be advised, however, that it may
take some time for NIOSH to update the currently available guide with the new PELs
Besides PELs and a wide variety of other valuable information, the pocket guide
includes the IDLH values described below
AIHA WEELs
The American Industrial Hygiene Association (AIHA) has established a committee to
develop Workplace Environmental Exposure Levels (WEELs) for toxic agents which have
no current exposure guidelines established by other organizations Essentially, the commit-
tee is attempting to establish occupational exposure limits for materials not addressed by the
ACGffiL or OSHA but of interest to various segments of industry A separate guide providing
documentation is being prepared for each substance
There are two WEEL limits for most materials The first is an 8-hour TWA value
similar in concept to ACGIH TLV-TWA values. The second, which is only available in a
limited number of cases, is a short-term TWA for exposures of either 1- or 15-minute
duration. As of October of 1988, WEELs were available for 33 materials Non-members
prices were $5 for each individual guide and $125 for the entire set (plus shipping and
handling).
The WEEL guides are available fiom AIHA Publications, 475 Wolf Ledges Parkway,
Akron, Ohio, 44311-1087 (Telephone- 216-762-7294) A price list and order form are
available at no charge
6-12
-------
NIOSHIDLHs
NIOSH defines Immediately Dangerous to Life or Health (IDLH) levels as the
maximum airborne contaminant concentrations "from which one could escape within 30
minutes without any escape-impairing symptoms or any irreversible health effects" Not
surprisingly, given that these limits are for 30-minute exposures under what are essentially
emergency conditions, IDLH values generally far exceed corresponding TLVs or PELs
They are available in the pocket guide referenced above for most substances currently
regulated by OSHA.
NAS/NRC EEGLs and SPEGLs
The Committee on Toxicology of the National Research Council (NRC), an operating
arm of the National Academy of Sciences (NAS), has published a list of Emergency
Exposure Guidance Limits (EEGLs) and Short-term Public Emergency Guidance Levels
(SPEGLs) as guidance in advance planning for the management of emergencies. Although
the Committee has been adding toxic substances to the list on a periodic basis, the careful
attention to detail and thoroughness of its work has resulted in EEGLs being established for
relatively few materials to date. Table 6.3 lists those available as of late 1988.
SPEGLs are concentrations whose occurrence is expected to be rare in the lifetime of
any one individual. These values, of which there are only four in the table, "reflect an
acceptance of the statistical likelihood of a nomncapacitating reversible effect in an exposed
population while avoiding significant decrements in performance". They are concentrations
considered acceptable for public exposures during emergencies
EEGLs differ from SPEGLs in that they are intended to apply to defined occupational
groups such as military or space personnel rather than the general public Because these
groups are typically younger and healthier, the EEGL for any particular substance may differ
substantially from the SPEGL.
Further information on these exposure limits and levels may be obtained by writing the
National Academy of Sciences, Committee on Toxicology, 2101 Constitution Avenue,
Washington, D.C, 20418 to the attention of Dr. Bakshi Note that the Committee plans to
have completed work on tnchloroethylene and lithium chromate by early 1989 if not sooner.
AIHA ERPGs
Several major chemical companies formed a task force in 1986 to develop Emergency
Response Planning Guidelines (ERPG) values for selected toxic materials The results of
their joint efforts are being published by the AIHA and are available from the publication
office cited earlier As of late 1988, guidelines had been completed for 10 substances
6-13
-------
TABLE 63
SUMMARY OF EMERGENCY EXPOSURE GUIDANCE LEVELS
FROM THE NATIONAL RESEARCH COUNCIL
Chemical
Acetone
Acrolein
Aluminum oxide
Ammonia
Arsine
Benzene
Bromotnfluoromethane
Carbon disulfide
Carbon monoxide
Chlorine
Chlorine tnfluonde
Chloroform
Dichlorodifluoromethane
(Freon-12)
60-MinuteEEGL
(ppm)
8,500
0.05
15 mg/m3
100
1.0
1000 (proposed)
25,000
50
400
3
1
100
10,000
Chemical
Dichlorofluoromethane
(Freon-21)
Dichlorotetrafluoromethane
(Freon-114)
1,1-Dimethylhydrazine
Ethanolamine
Ethylene oxide
Ethylene glycol
Fluonne
Hydrazine
Hydrogen chlonde
Hydrogen chlonde
Hydrogen sulfide
Isopropyl alcohol
Lithium bromide
60-MinuteEEGL
(ppm)
100
10,000
024*
50
20 (proposed)
40
7.5
012
20
1*
10(24hr)
400
15 mg/m3
ON
I-1
-P-
Note' Units in parts per million by volume in air unless otherwise stated.
*SPEGL (Short-term Public Emergency Guidance Levels)
11/88
-------
TABLE 6.3 (Continued)
SUMMARY OF EMERGENCY EXPOSURE GUIDANCE LEVELS
FROM THE NATIONAL RESEARCH COUNCIL
Chemical
Mercury vapor
Methane
Methanol
Monomethyl hydrazme
Nitrogen dioxide
Nitrous oxide
Ozone
Phosgene
60-Minute EEGL
(ppm)
02mg/m3(24hr)
5,000 (24k)
200
024*
1*
10,000
1
02
Chemical
Sodium Hydroxide
Sulfur dioxide
Sulfunc acid
Toluene
Tnchlorofluoromethane
(Freon-11)
Tnchlorotrifluoroethane
(Freon-113)
Vinyhdene chlonde
Xylene
60-Minute EEGL
(ppm)
2mg/m3
10
1 mg/m3
200
1,500
1,500
10(24hr)
200
H
Ul
Note: Units in parts per million by volume in air unless otherwise stated
*SPEGL (Short-term Public Emergency Guidance Levels)
11/88
-------
including ammonia, chlorine, chloroacetyl chloride, chloropicrm, crotonaldehyde, diketene,
formaldelhyde, hydrogen fluoride, perfluoroisobutylene, and phosphorous pentoxide. Pub-
lished in two sets of five, the first set costs $7 while the second is pnced at $11.
As in the case of NAS/NRC efforts, the task force is attempting to define toxic
exposure limits suitable for use in advance planning for emergencies. It ultimately wishes,
however, to address a much greater number of chemicals than those considered to date by the
NAS/NRC.
The task force intends to establish thiee limits for each material, these being:
• ERPG-3: The maximum airborne concentration below wjiich, it is be-
lieved, nearly all individuals could be exposed for up to one hour without
experiencing or developing life threatening health effects.
• ERPG-2: The maximum airborne concentration below which, it is be-
lieved, nearly all individuals could be exposed for up to one hour with out
experiencing or developing irreversible adverse or other senous health
effects or symptoms which could impair an individual's ability to take
protective action. This particular limit is being developed using criteria
similiar to those applied by the NAS/NRC.
• ERPG-1: The maximum aiiborne concentration to which nearly all
individuals could be exposed for up to one hour without experiencing or
developing health effects more severe than sensory perception or mild
irritation, if relevant.
6.8 ADVANTAGES AND DISADVANTAGES OF VARIOUS LIMITS
A key problem of using TLV, PEL, or WEEL values in the course of evacuation
planning or hazard assessment is that they are intended for use in the occupational
environment where presumably healthy workers are exposed to concentrations near these
limits day after day throughout their careers This, and the desire to prevent health effects
associated with both acute and chronic exposures, means that these values are often (but not
always) much lower than what they need be to protect the public from exposures associated
with rare or infrequent spills of brief duration. Consequently, use of a TLV, PEL, or WEEL
value, although decidedly safe in the vast majority of cases, could conceivably result in major
overprediction of downwind evacuation or hazard zones in many situations Key exceptions
involve materials such as chlorine, acids, caustics, and other generally corrosive materials for
which limits are based on irritant rathei than toxic effects and for which applied safety
factors may be minimal.
6-16
-------
NIOSH IDLH limits are considerably higher, are defined for an exposure duration
closer to what would be expected in many actual short-term spill emergencies, and are closer
to the borderline between levels that are barely tolerable and those that may cause significant
injury The problem is that "barely tolerable" contaminant concentrations may have the
potential to cause considerable irritation or other distress, possibly to the point of prompting
large numbers of people to seek medical assistance. Also, since NIOSH is again assuming
that healthy workers are being exposed, IDLH concentrations may have the potential to cause
significant injury to young children, the elderly, or individuals with preexisting health
problems. Consequently, it is apparent that a safety factor must be applied if the IDLH is
used in any way for protection of the general public, especially if exposures exceed 30
minutes in duration
The NAS/NRC SPEGLSs and AfflA ERPG-2 values are possibly the best choice
among currently available guidelines for protection of the public during relatively short-term
events given the objectives of their respective developers. Unfortunately, only a small
number of hazardous materials have been addressed to date.
Overall, the above discussion might seem to suggest there is no widely accepted
method available for selection of an appropriate exposure limit for general populations
subjected to toxic vapors or gases, particularly where the exposure limit is to be used for
public emergency planning purposes That is indeed (and unfortunately) an accurate
appraisal of the current situation. So what should you do? Some options, in order of
decreasing preference, and by no means mandatory for use, are as follows:
Use the NAS/NRC SPEGL or the AfflA ERPG-2 value for the material if
one has been established
• Consult a lexicologist or similarly qualified individual for advice based on
a formal review of the toxicity of the material of concern
• Use the highest value among the following:
IDLH value divided by 10 (with " 10" being a safety factor)
TLV-STEL
TLV-TWA multiplied by 3 (if a TLV-STEL does not exist)
TLV-C
• If the evacuation of additional areas is not a problem, or the exposure may be
prolonged beyond one hour, use the TLV-TWA or the TLV-C value or apply an
additional safety factor to other selections
6-17
-------
The above suggestions should not be considered more than rough guidelines that will
generally lead to an adequately "safe" answer for most members of a community. There is,
however, one more problem to consider.
The chronic exposure limits for substances known or suspected to be carcinogens are
usually set at very low levels to protect workers from developing cancer during their
lifetimes. Such values are generally many times (possibly several hundred times) lower than
the limits enforced for the same materials prior to the discovery of a potential cancer threat.
For example, the TLV-TWA for vinyl chloride is now 5 ppm whereas it was 200 ppm for
many years, yet even 200 ppm is well below any concentration causing observable health
effects in short-term acute exposures. Obviously, the size of the evacuation or hazard zone
for a 5 ppm limit would be many times larger than a zone with boundaries of 200 ppm. The
difference in the numbers of people that may require evacuation or other protective action
may differ by thousands if not tens of thousands in urban areas.
There is no hard evidence that a single exposure to a substance such as vinyl chloride
will cause excess cancers in a population of exposed humans However, some scientists are
of the opinion that any exposure might lead to at least a minor increased nsk of such cancers,
and this belief poses a dilemma during planning for evacuations, especially given the public
fears that may naturally accompany the announcement that a cancer-causing agent has been
released into the atmosphere It is therefore necessary for government and industry to
consider cases involving carcinogens carefully and on a case-by-case basis, giving full
attention to the safety issues involving large-scale evacuations as well as the potential
long-term health, political, and legal implications of their decisions.
6.9 RELATIONSHIP OF RECOMMENDATIONS TO EPA LOCs
Under the Superfund Amendments and Reauthonzation (SARA) Act of 1986, the U S
Environmental Protection Agency (EPA) established a list of several hundred Extremely
Hazardous Substances (EHS) subject to emergency planning, community right-to-know,
hazardous emissions reporting, and emergency notification requirements. In providing
guidance to planning personnel for screening and prioritizing threats posed by EHS, the EPA
made a first attempt at specifying what it termed Levels of Concern (LOCs) for these
substances, essentially adopting portions of the approach recommended above.
For the 390 or so substances for which NIOSH has established IDLH levels, the EPA
set LOCs to one-tenth of available IDLHs until such time as industry and government
develop more appropnate exposure limits for protection of the public during episodic
short-term emergencies For substances for which IDLHs had not been established, the EPA
developed a highly approximate procedure to estimate LOCs comparable to IDLHs
Essentially, IDLHs were estimated for new substances via use of data obtained from
6-18
-------
laboratory experiments involving acute exposures of animals to toxic substances. Inhalation
data were used in preference to data for other routes of exposure when available, but data for
other routes of exposure were indeed used when necessary The following equations were
then applied to convert available data to airborne concentrations presumably comparable to
IDLHs.
1) Estimated IDLH = LC^ x 0.1
2) Estimated IDLH = LCLo
3) Estimated IDLH = LD^x 0.01
4) Estimated IDLH = LDLo x 0.1
The abbreviations used above for lethal concentrations and dosages are defined and
described in Section 6.12 of this chapter. Please note that the above discussion only provides
a general overview of the EPA's general approach and should not be applied in an
indiscriminate fashion.
6.10 CONSIDERATION OF MIXTURES OF HARMFUL GASES AND VAPORS
Preceding discussions have focused on relatively pure substances. As is well appreciat-
ed, however, many materials handled by industry are multi-component mixtures. It is well
therefore to discuss how to determine appropriate toxic limits for mixtures via a review of
traditional guidance found in the literature.
The ACGIH, in an appendix to its TLV booklet, reports that one of the first tasks in
looking at mixtures is a determination of whether mixture components have additive or
independent effects on the human body. In other words, when two or more toxic agents in a
mixture act upon the same organ system, it is their combined or additive effect rather than
their individual effects that should be given primary consideration, and indeed, this is the
preferred approach in the absence of specific information to the contrary. Where toxicologi-
cal data firmly support a finding that the chief effects of the different substances are not in
fact additive (in the sense that they produce purely local effects or affect different organs of
the body), it is only then acceptable to assume that adverse effects are independent.
Where effects are evaluated as being additive, the ACGIH suggests that the sum of the
following fraction be computed:
Ci C2 Cn
Sum=-+-...-
6-19
-------
where: Cn indicates the measured or predicted atmospheric concentration, Tn
indicates the corresponding toxic limit in the same units as Cn, and
there are "n" number of toxic substances in the mixture.
When the Sum of the fractions equals 1.0 or less, then the vapor mixture is considered
to be at or below the toxic limit In those cases where all components of a mixture are
deemed to produce independent effects, the toxic limit is considered to be exceeded only
when one or more of the individual CJ Tn fractions has a value greater than one.
To be noted is that synergistic action or potentiation may occur with some combina-
tions of toxic agents: these being cases in which the combined effect of the mixture actually
exceeds the impact indicated by assumption of additive effects Such cases, which are
fortunately rather rare, must be considered on a case-by-case basis.
When the source of airborne contamination is a liquid mixture, the ACGIH suggests (to
its typical audience of industrial hygiemsts and other occupational health personnel) that the
composition of the airborne mixture be assumed similar to the composition of the original
liquid mixture. In effect, this results in the further assumption that all components of the
mixture will evaporate at a constant rate in direct proportion to then* concentration in the
liquid mixture. The assumption has ment when one in interested in evaluation of a relatively
long-term time-weighted average exposure resulting from a mixture that will eventually
evaporate in its entirety, but has severe limitations when applied to the assessment of acute
exposures resulting from accidental and episodic events. It is well, nevertheless, to present
the ACGIH's general methodology for estimating the toxic limit of a liquid mixture of this
type, this being:
Toxic Limit (mixture) = „—•=.—=-
£j_, £2 £»
Ci <*'"<*
where: Fn indicates the weight fractions of individual components in the liquid
mixture, and Cn indicates the corresponding toxic limits in units of mg/m3.
A more formal approach to determining the airborne mixture toxic limit for evaporat-
ing or boiling pools of liquid requires consideration of vapor-equihbnum factors beyond the
scope of this text. Nevertheless, where needs for a more precise limit are critical, it is
desirable to apply more sophisticated analytical procedures to evaluate vapor compositions
above liquid mixtures or to make direct measurements of representative samples The
procedures for such efforts are well within the state of the art of engineering practice and
entail fundamental principles of thermodynamics
6-20
-------
6.11 EXPOSURE LIMITS FOR CONTAMINATED WATER
The U.S. Environmental Protection Agency (EPA) has established or recommended
water quality criteria for a variety of water uses and a relatively large number of chemicals.
Advice from their personnel as to what concentrations of any particular chemical are
tolerable in any given situation may be available with only a telephone call to one of the
EPA's 10 regional offices
Among the various standards and guidelines developed by the EPA for protection of
water quality are:
• National Drinking Water Standards The maximum contaminant levels
(MCLs) for selected heavy metals, pesticides, radioactive substances, and
other water quality characteristics permitted by law in water destined for
human consumption Listed in Parts 141 and 143 of Title 40 of the Code of
Federal Regulations (CFR).
• Drinking Water Health Advisones (HAs) -- previously called Suggested No
Adverse Response Levels (SNARLS): Human health effects advisories for
unregulated drinking water contaminants commonly found in potable water
supplies. HAs are somewhat unique in that they provide guidance for
short-term exposure as well as the long-term chronic exposures typically of
interest to the EPA.
• Maximum Contaminant Level Goals (MCLGs) — formerly known as
Recommended Maximum Contaminant Levels (RMCLs). Published in the
Federal Register of June 12, 1984, the EPA proposed zero contamination
limits for six halogenated hydrocarbons and benzene. Low levels of
contamination were permitted for two other halogenated hydrocarbons (i e,
1,1,1-tnchloroethane and 1,4-dichlorobenzene) MCLGs were recently
proposed for several additional contaminants.
• Federal Water Quality Criteria Criteria for acute and chronic exposure of
freshwater and saltwater aquatic life and human health based on long-term
consumption of drinking water and contaminated fish or shellfish. Avail-
able for a relatively long list of substances.
Spills of toxic materials into a body of surface water differ from discharges of toxic
vapors or gases into the an* in that a large number of people are unlikely to suffer toxic
effects before authorities have a chance to restrict water use Indeed, response planning for
the spill of any hazardous material into water more typically involves preparations to:
6-21
-------
• Alert proper state and federal authorities.
• Warn public, industrial, agricultural, and recreational users of the water on
as prompt a basis as possible of the contamination
• Attempt to limit the degree of contamination or the amount of water
affected.
• Attempt to remove as much of the contaminating substance as possible
from the water (possibly employing a contractor with specialized expertise
and equipment).
• Analyze the water to determine the extent of contamination.
• Consult with proper authorities as to whether the water is fit for use or
whether other remedial actions are first necessary; and
• Prepare for the eventuality that a particular water supply may become
unavailable for use for a time.
6.12 UNDERSTANDING TOXICOLOGICAL DATA IN THE LITERATURE
Toxicologists have a number of "short-hand" methods of expressing the toxicity of
hazardous materials by various routes of entry. An understanding of some of the more
common abbreviations used can lead to a greater understanding of how the toxicities of
various materials can be assessed, particularly when these abbreviations are encountered in
hazardous material data bases or the safety related literature of chemical manufacturers that
address the effects of acute exposures.
The easiest way to learn the abbreviations is to look at a few examples and then discuss
their meaning:
The orl rat LDj,, for Chemical A is 200 mg/kg.
• The ihl LCj,, for the mus or gpg is 800 ppm/4 hrs The TCLo is 100 ppm/4
hrs.
• The rbt skn LDCT is 50 mg/kg.
LD in the above examples is an abbreviation for lethal dose while LC stands for lethal
concentration. TC is short for toxic concentration while TD means toxic dose. There are
other similar abbreviations but these are by far the most common.
6-22
-------
Each of the LD or LC notations are followed by a number that is usually a subscript A
"50" means that 50% of the test population of animals were killed under stated test
conditions, a "67" means 67% were killed and so forth. The letters "Lo" instead of a number
mean this is the lowest reported level having the stated effect
In order for one of the above notations to have meaning, both the species of animal
tested and the route of entry must be specified. Typical abbreviations are:
Species of Animal
Rat = rat
Mouse = mus
Guinea pig = gpg
Rabbit = rbt
Human = hmn
Mammal = mam
Monkey = mky
Route of Entry
Oral = orl
Skin application = skn
Inhalation = ihl
Both oral and skin application dosages are typically expressed in units of milligrams of
chemical applied per kilogram of the animal's body weight, or mg/kg for short The actual
total amount of a toxic material necessary to cause the stated effect is determined by
multiplying the dose in units of mg/kg by the weight of the animal species expressed in units
of kilograms (1 kg = approximately 2 2 Ib).
Inhalation data must include the concentration in air to which the animal species was
subjected as well as the duration of exposure. Concentrations in air are typically expressed in
units of ppm (by volume) or mg/m3 Times are typically given in minutes or hours Be
advised that any airborne concentration not accompanied by an indication of the duration
of exposure should be considered a useless and thoroughly meaningless item of informa-
tion
One of the most comprehensive compilations of lexicological data is a multi-volume
set of documents tided Registry of Toxic Effects of Chemical Substances The 1985-1986
edition, published in April 1987, is available for a cost close to $100 from the Superintendent
of Documents, U.S Government Printing Office, Washington, D C, 20402, or one of the
6-23
-------
many regional offices of the GPO, as Stock No 17-33-00431-5. Developed jointly by the
U.S. Public Health Service, Centers for Disease Control and NIOSH, the set is also listed as
DHHS (NIOSH) Publication No. 87-114.
6-24
-------
7.0 REACTIVITY HAZARDS OF CHEMICAL SUBSTANCES
7.1 INTRODUCTION
It has up to this point been assumed that the hazardous materials being discharged or
spilled do not in any way react with or chemically transform due to contact with water, an-,
other common materials in the environment, or other chemicals that may be present in the
vicinity It has also assumed that these materials are not self-reactive under conditions that
may be encountered. Although the overall topic of chemical reactivity hazards is extremely
complex, it is necessary to at least briefly outline some of the more common and/or
dangerous types of reactions and how they may pose a threat to nearby populations With
due apologies to chemists, chemical engineers, and others with a knowledge of these topics,
it is acknowledged that some liberties are taken in this process to ensure that various
concepts are more easily understood by non-technical audiences.
73 EXOTHERMIC REACTIONS
When one substance is brought together or mixed with another and the resulting
interaction evolves or generates heat, the process is referred to as an exothermic reaction.
Alternatively, if no reaction will take place unless heat is continuously added to the
combination of reactants, the interaction resulting from the provision of heat is called an
endothermic reaction. However, it is important to understand that some exothermic
reactions may require heating just to get started, and will then proceed on their own
-------
Exothermic reactions pose special hazards whether occurring in the open environment
or within a closed container. In the open, the heat evolved will raise the temperature of the
reactants, of any products of the reaction, and of surrounding materials. Since several
properties of all substances are a function of temperature, the resulting higher temperatures
may affect how the materials involved may behave in the environment. Of key importance is
the realization that heat will increase the vapor pressures of hazardous materials and the rate
at which they vaporize. If very high temperatures are achieved, nearby combustible
materials may ignite. Explosive materials, be they the reactants or products of the reaction,
may explode upon ignition or excessive heating
Similar hazards are associated with exothermic reactions taking place in closed
containers. In this case, however, increasing internal temperatures as well as the evolution of
gases from the reaction may increase internal pressures to the point that the tank or container
ruptures violently in an overpressunzation explosion, thus suddenly releasing large amounts
of possibly flammable and/or toxic gases or vapors into the atmosphere Such gases or
vapors may also be released through ruptured pipes, opened pressure relief devices, or any
other paths to the external environment
Reactions with Water or Air
Some of the most basic types of exothermic reactions (which are barely "reactions" in
the true sense of the term) occur when certain materials are dissolved in water Such
substances have what is called a positive heat of solution. They do not transform to a
different material, but simply generate heat while mixing Some examples are sodium
hydroxide (also called caustic soda) and sulfunc acid, which generates considerable heat to
the point of causing some degree of "violence" when concentrated or pure materials are
spilled into water. Yet other materials may ignite, evolve flammable gases, or otherwise
react violently when in contact with water. Knowledge of the reactivity of any substance
with water is especially important when water is present in the spill area or a fire takes place
and firefighters do not wish to make the situation worse by applying water to the flames or
chemicals.
While discussing such substances, it is well to add that several of the strong acids and
related substances in this category of materials may evolve large amounts of fumes when in
contact with water or moisture in the air. These fumes, which may consist of a mixture of
fine droplets of acid in air and acid vapors, are usually highly irritating, corrosive, and
heavier than air.
Many substances referred to as being pyrophoric will react violently or exothermically
with air and are likely to ignite in a spontaneous fashion. Such substances (such as
phosphorus) are commonly transported or stored in a manner that prevents exposure to air,
7-2
-------
often submerged in water or some type of compatible oil. Note that the fact that a substance
can be safely stored under water in no way suggests that it may also be safely submerged in
oil. Nor may submersion in water be safe for a substance usually maintained under some
type of oil.
Reactions with Combustible Organic Materials
Certain chemicals are known as strong oxidizing agents or oxidizers. They have the
common characteristic of being able to decompose or oxidize organic materials and react
with a variety of inorganic materials while generating heat, oxygen, flammable gases, and
possibly toxic gases. If the heat generated is sufficient to ignite a combustible or flammable
material, a fire or explosion may occur.
Another group of chemicals are referred to as strong reducing agents. These
substances may evolve hydrogen upon reaction with many other chemicals, may evolve other
flammable or toxic gases, and like oxidizing agents, may generate heat. As above, a fire or
explosion may result if sufficient heat is generated to ignite a combustible or flammable
substance Strong reducing agents and oxidizing agents should never be allowed to make
contact without appropriate safeguards since they represent opposite extremes of chemical
reactivity.
Exothermic Polymerization Reactions
A few of the more common plastics in use on a widespread basis are polyethylene,
polypropylene, polystyrene, and polyvinyl chloride (PVC). Although all are manufactured
from liquids or gases, they are typically solids in then- final form
The above plastics are respectively manufactured from ethylene, propylene, styrene,
and vinyl chloride by means of a polymerization reaction in which molecules of these
materials are linked together into long chains of molecules. As the chains become longer and
begin connecting to each other, thus greatly increasing the molecular weight of individual
molecules, a solid plastic is formed
Some chemicals capable of being polymerized have a strong tendency to do so even
under normal ambient conditions and are especially prone to polymerize if heated above a
certain temperature or if contaminated by a catalyst or polymerization initiator, which in
some cases might be a rather common substance such as water or rust. Once polymerization
starts, an exothermic chain reaction may occur that develops high temperatures and pressures
within containers and which can lead to possible explosion or violent rupture of the container
and/or discharge of flammable and/or toxic gases if safety and control systems malfunction
or are lacking The incident in Bhopal, India partially involved this type of reaction when a
container of methyl isocyanate contaminated with water and chloroform began polymerizing.
7-3
-------
The heat of the runaway (i.e., out of control) reaction caused a large portion of the highly
toxic isocyanate to vaporize into the air through a pressure relief system before it had a
chance to polymerize.
Quite often, substances with the above tendency to self-polymerize or to undergo
autocatalytic polymerization are transported or stored only while containing an amount of a
substance called an inhibitor. As their name implies, inhibitors act to inhibit, slow, or
interfere with the chemical processes that can lead to a runaway uncontrolled polymerization
reaction under normal conditions of transportation or storage. Inadvertent contamination or
excessive heat, however, may overpower the inhibitor and allow the reaction to proceed
Thus, an inhibited cargo should not be considered safe if there is a possibility of it being
overheated or contaminated with those substances that may initiate polymerization. The very
fact that a substance needs an inhibitor for safe storage is in many cases (but by no means all)
a sign of potential hazardous instability.
Exothermic Decomposition Reactions
Much as some chemical molecules can join together to form larger molecules via
exothermic polymerization, others are unstable and can break apart in a runaway exothermic
reaction once the process is initiated Again, inhibitors may be used to slow the process
down or to prevent its occurrence and various contaminants or heat may overcome the
inhibitors or otherwise start a reaction. Containers may explode, rupture, and/or vent various
flammable and/or toxic gases to the atmosphere
Incidentally, the above decomposition and polymerization reactions are hazardous only
if they somehow become uncontrolled and start a chain reaction that cannot be stopped with
available equipment, materials, or safety systems. They are widely and safely conducted in
chemical and other manufacturing plants across the nation on a daily basis without incident
It is only when control or safety systems break down or people make mistakes that problems
begin.
73 NEUTRALIZATION REACTIONS
Spill response guides often suggest consideration of neutralization as a way in which a
hazardous substance can be converted via a chemical reaction to one or more substances that
pose lesser threats to the public health or the environment It is therefore worthwhile to say a
few words on the topic
In the traditional sense of the word, neutralization typically refers to the combination of
an acid and a base or alkaline material to form some sort of salt A good example involves
the careful combination of sodium hydroxide (caustic soda — NaOH) with hydrochloric acid
(muriatic acid -- HC1 in water). This reaction, which may proceed violently for a time,
7-4
-------
generate heat and gases, and cause boiling and spattering of acid if not properly controlled,
results in the combination of sodium (Na) atoms with chlorine (Cl) atoms to form sodium
chloride (NaCl), which is ordinary table salt The remaining hydrogen (H) atoms and
hydroxide (OH) molecules combine to form ordinary water (Hp). Thus, one strongly
corrosive and hazardous substance is used to convert another to a solution of ordinary salt
and water.
When used in the spill response community, neutralization refers to the general use of
one or more chemicals or other substances to render another less harmful The term need not
solely apply to acid-base reactions.
7.4 CORROSIVITY HAZARDS
The process by which a chemical gradually erodes or dissolves another material is
often referred to as corrosion. The process represents yet another type of chemical reactivity
that must be considered in assessing the hazards of any given material, and is particularly
important when. 1) choosing materials of construction for container walls or linings, piping,
pumps, valves, seals, gaskets, and so forth; and 2) assuring that equipment and materials used
during response to emergencies will not be damaged or destroyed by contact with the spilled
material during their period of use. The word corrosive is also used descriptively to indicate
that a substance may cause chemical burns of the skin, eyes, or other bodily tissues upon
contact
In evaluating whether one material is corrosive to another via reference to material
safety data sheets, chemical company product bulletins, hazardous material data bases, or
other reference sources, it is often important to place the time frame and rate of corrosion
into the proper context For example, certain reference sources may state that one substance
is unacceptably corrosive to a particular material of construction because long term (i.e., 10
to 20 years) exposure will result in failure of the material prior to the desired lifetime of the
equipment Yet other reference sources may discuss the issue in terms of short term
resistance of equipment or clothing construction materials to chemical attack, particularly if
addressing use under emergency conditions This distinction is not always clear in the
literature
Finally, note that some of the most corrosive substances to common metals include
strong acids of one type or another Not only may the "wrong" acid in contact with the
"wrong" metal cause rapid corrosion of the metal, but the process may generate flammable
and potentially explosive hydrogen gas
7-5
-------
73 OTHER HAZARDOUS RESULTS OR PRODUCTS OF REACTIONS
The above discussions have really only scratched the surface of the overall topic of
hazardous chemical reactions. It is also necessary to point out that:
• The combination of various chemicals may produce new chemicals with
hazards quite different and possibly more severe than those associated with
the original materials.
• Some combinations may result in spontaneous fires; spontaneous explo-
sions; formation of substances which will ignite or explode if shocked,
heated or subjected to friction; generation of toxic gases, liquids or solids;
or generation of flammable gases, liquids, or solids
• It is necessary to look at hazardous materials on a fairly specific
case-by-case basis to determine their reactivity hazards.
7.6 SOURCES OF CHEMICAL REACTIVITY DATA
There are numerous sources of chemical reactivity data that address the topic
somewhat superficially and several that are highly technical and somewhat beyond the
perceived "needs" of the audience to which this document is directed. The following three
sources provide an excellent balance between completeness, precision, specificity, and
common availability.
• Fire Protection Guide on Hazardous Materials, containing "Manual of
Hazardous Chemical Reactions," NFPA 491M-1986, National Fire Protec-
tion Association, Batterymarch Park, Qumcy, MA 02269 (Telephone
1-800-344-3555 for orders).
• Brethenck, L., Handbook of Reactive Chemical Hazards, 3rd edition,
Butterworths, London and Boston, 1985. Available through libraries and
bookstores serving the scientific community
• Hatayama, H K., et al, A Method for Determining the Compatibility of
Hazardous Wastes, EPA Report No. EPA-600/2-80-076, Municipal Envi-
ronmental Research Laboratory, U.S. Environmental Protection Agency,
Cincinnati, Ohio, Apnl 1980. Available as publication PB80-221005 from
the National Technical Information Service, Springfield, Virginia 22161.
7-6
-------
The NFPA Fire Protection Guide on Hazardous Materials contains the described
manual of chemical reactions as well as considerable additional information and data on
hazardous materials. Found in the libraries of numerous fire departments, it was available in
1988 at a cost of approximately $49 to non-members. Although the section on hazardous
chemical reactions has not truly been updated since 1975, and is not nearly as extensive as
the work by Brethenck, the guide remains an excellent source for a broad range of specific
information. Major sections of the guide can also be found in the NFPA National Fire Codes
as Sections 325M, 49,491M, and 704.
The handbook by Brethenck covers approximately 9000 compounds versus the
1600-1700 found in the NFPA guide. It is a major and somewhat unique work in the field
which retails for $110.
The report prepared by Hatayama and his co-workers under the sponsorship of the EPA
is an excellent supplement to either of the above sources of information. Those above mostly
list and describe the specific hazardous consequences of combining various sets of
chemicals, as reported in the general literature Since there are many tens of thousands of
known chemicals, and since only a small fraction of the possible combinations have been
reported upon, neither of these sources can even begin to claim that combinations not listed
are safe. The work by Hatayama et al. attempts to fill the gaps by providing a general
indication of the typical effects of mixing a substance from one chemical family with a
substance from another family via a single chemical compatibility chart. The title of the
work suggests it only considers hazaidous waste materials, but that in no way affects the
validity of the information for hazardous materials in general Appendix D to this guide
contains a copy of the chart as well as additional explanatory information.
It is also necessary to note that many of the product bulletins and safety-related
documents available free from most chemical manufacturers can be excellent sources of
information when one is concerned with the reactivity hazards of a relatively small number
of materials. The problem is that collection of such publications for a large number of
materials can be a burdensome and lengthy process.
Chemical company literature, however, can be a great source of information on the
compatibility of common materials of equipment construction with specific chemicals.
Alternatives include some of the better hazardous materials data bases, books devoted to this
topic, and more widely available handbooks in the fields of chemical and mechanical
engineering. Many of these same sources address the compatibility of materials used for
chemical protective clothing, and substantial information is available from the manufacturers
of such clothing One excellent source of information on protective clothing that deserves
special notice is:
7-7
-------
Schwope, A D., et al, Guidelines for the Selection of Chemical Protective
Clothing, 3rd edition, 1987; sponsored by the EPA and USCG and
available for approximately $35 from the ACGIH Publications Section,
6500 Glenway Ave, Bldg D-7, Cincinnati, Ohio 45211 (Telephone
513-661-7881).
7-8
-------
8.0 HAZARDOUS MATERIAL CLASSIFICATION SYSTEMS
5
B-
\
111
00
NFPA
J INTRODUCTION
Vanous organizations in the United States have established or defined classes or lists of
hazardous materials for regulatory purposes or for the purpose of providing rapid indication
of the hazards associated with individual substances. An awareness and knowledge of these
classification systems can assist emergency preparedness personnel in identifying those
materials that may pose a potential threat to their respective jurisdictions.
8.2 U.S. DEPARTMENT OF TRANSPORTATION CLASSIFICATIONS
As the primary regulatory agency concerned with the safe transportation of hazardous
materials in interstate commerce, the US Department of Transportation (DOT) has
established definitions of various classes of hazardous materials, established placarding and
marking requirements for containers and packages, and adopted an international cargo
commodity numbering system. Each of these topics is individually discussed in the
following Further details are available in 49 CFR 171-179.
Material Classification Definitions
The DOT classifies hazardous materials in transportation into one or more of the
following categories.
-------
An explosive is defined as "any chemical compound, mixture, or device, the primary or
common purpose of-which is to function by explosion, i e., with substantially instantaneous
release of gas and heat. " within certain limitations noted in DOT regulations The overall
category of explosives is further broken down into Class A, Class B, and Class C explosives
Class A materials are among the most powerful and include bombs, mines, torpedoes, and
ammunition used by the military; various high explosives like nitroglycerm and dynamite;
blasting caps, detonating fuzes, and powerful rocket propellants. Class B substances and
devices are generally less powerful and typically (not always) function by rapid combustion
rather than detonation. The class includes special fireworks, flash powders, some pyrotech-
nic signal devices, liquid or solid propellants, some smokeless powders, and certain types of
ammunition. Class C explosives are manufactured articles which contain Class A or B
materials, or both, as components in strictly restricted quantities The class also includes
certain types of fireworks
A blasting agent is a material designed for blasting which has been tested in
accordance with DOT regulations and "found to be so insensitive that there is very little
probability of accidental initiation to explosion or of transition from deflagration to
detonation" In other words, the material is capable of exploding under very special
conditions, but these conditions are unlikely to occur in transportation, even in the event of
an accident involving fire or impact
Flammable liquid refers to any liquid, within certain limitations and exceptions, that
has a "closed-cup" flash point below 1GO°F (37 8°C) Similarly, combustible liquid refers to
any liquid that has a flash point of 100°F or more but no higher than 200°F A pyrophoric
liquid is any liquid that ignites spontaneously in dry or moist air at or below 130°F (54 5°C)
Flammable solids are "any solid material, other than one classed as an explosive,
which, under conditions normally incident to transportation is liable to cause fires through
friction, retained heat from manufacturing or processing, or which can be ignited readily
and -when ignited burn so vigorously and persistently as to create a serious transportation
hazard Included in this class are spontaneously combustible and water-reactive materials "
An oxidizer, according to DOT regulations, "is a substance such as a chlorate,
permanganate, inorganic peroxide, or a nitrate, that yields oxygen readily to stimulate the
combustion of organic matter" The key hazard associated with oxidizing agents or materials
is that contact with a combustible substance, particularly organic materials, may cause the
substance to ignite and possibly even explode
An organic peroxide is essentially a derivative of hydrogen peroxide (H/)^ where one
or more of the hydrogen atoms have been replaced by molecular chains containing carbon
8-2
-------
and hydrogen atoms. The substances in this category do not meet the definitions of Class A
or B explosives but may be capable of exploding under certain conditions They may also
have the hazards associated with oxidizers.
DOT defines a corrosive material as "a liquid or solid that causes visible destruction or
irreversible alterations in human skin tissue at the site of contact, or in the case of leakage
from its packaging, a liquid that has a severe corrosion rate on steel" A liquid is considered
to have a severe corrosion rate if it "eats away" more than 0.25 inch of a certain type of steel
at 130°F over the course of one year.
A compressed gas is defined as any material or mixture with an absolute pressure in a
container of
More than 40 psia at 70°F
More than 104 psia at 130°F
• If the substance is flammable and in the liquid state, more than 40 psia at
100°F.
A flammable compressed gas is a compressed gas that has a lower flammable limit (LFL)
concentration of 13% or less by volume in air, or which has a flammable range (i e, the
difference between the LFL and UFL) of greater than 12%, or which behaves in a
prespecified manner in a flammability testing apparatus. A liquefied compressed gas is a gas
which is partially a liquid under the pressure in the container at 70°F. A non-liquefied
compressed gas is a substance which is entirely gaseous at a temperature of 70°F.
Poisonous materials are divided into three groups in DOT regulations according to their
degree of hazard in transportation. Poison A substances are "poisonous gases or liquids of
such a nature that a very small amount of the gas, or vapor of the liquid, mixed with air is
dangerous to life " Poison B materials are liquids or solids, other than Class A poisons or
irritating materials, "which are known to be so toxic to man as to afford a hazard to health
during transportation, or which, in the absence of adequate data on human toxicity, are
presumed to be toxic to man" because they meet certain criteria for inhalation, ingestion, or
skin exposures when tested on laboratory animals Irritating materials are liquid or solid
substances "which upon contact with fire or when exposed to air give off dangerous or
intensely irritating fumes"
In the aftermath of the Bhopal incident, DOT rules were modified to require special
marking of packages or containers of volatile toxic liquids which had previously escaped
classification as poisons After adopting a new set of special criteria for inhalation toxicity
hazards, the DOT required that packages containing more than one liter and no more than
8-3
-------
110 gallons of these materials be marked Inhalation Hazard. Poison placards were required
in addition to other required placards for trucks, rail cars or containers carrying any amount
of these materials. Shipping papers for containers holding more than one liter were required
to include the statement Poison - Inhalation Hazard.
An etiologic agent is "a viable microorganism, or its toxin, which causes or may cause
human disease " For the most part, such agents include potentially infected living tissue and
bacteriological materials.
Radioactive materials are substances that give off potentially harmful nuclear
radiation, and are classed in three gioups according to the controls needed to provide
"nuclear criticality safety" during transportation. Fissile Class I materials are among the
safest of these substances, do not require nuclear cnticahty safety controls dunng transporta-
tion, and may be shipped together in an unlimited number of packages Fissile Class II
substances are somewhat more dangerous and can only be shipped in limited amounts when
packages are shipped together. Fissile Class HI do not meet the requirements of the other
classes and must be controlled to provide nuclear criticality safety in transportation by
special arrangement between the shipper and the earner
Finally the DOT has a category called Other Regulated Material (ORM) for a wide
variety of hazardous materials shipped in limited quantities and in certain kinds of packaging
There are five classes of such cargos with the designations ORM-A, ORM-B, ORM-C,
ORM-D,andORM-E
Identification Numbers
The DOT has assigned a four-digit identification to each of the hazardous materials
regulated in transportation. When appearing in documentation, these numbers are preceded
by the letters "UN" or "NA" The UN numbers, such as UN1203 for gasoline, were assigned
in cooperation with the United Nations and are used on an international basis The NA
numbers are not recognized in international transportation except to and from Canada.
Most of the numbers and the material shipping names to which they are assigned
represent very specific materials. It is well to recognize, however, that the DOT also permits
some cargos to be identified in a rather genenc fashion. For example, the identification
number UN1993 applies to flammable liquid, nos The last three letters are an abbreviation
for not otherwise specified, so the number does not permit identification of the specific
material in the container.
8-4
-------
Placards and Labels
The DOT has numerous regulations dealing with the placards and labels that must
appear respectively on bulk containers and packages of hazardous materials. Figure 8.1
illustrates the required placards, these being the fairly large signs that must appear on railroad
tankcars, highway tank trucks, and other large transport vehicles Labels are fairly similar
and any differences are rather self-explanatory.
Special Notes
Before continuing, it is necessary to make two important observations about DOT
classification systems and placarding and labeling requirements The first is that these
systems and requirements are modified on a frequent basis and there has been considerable
activity to improve them in the aftermath of the Bhopal incident. Although the material
presented herein is of a fairly general nature, some items may become outdated with time.
Indeed, even as this document was being prepared, the DOT was in the process of finalizing
new regulations in this area.
Secondly, and most importantly, be intensely aware that the current DOT material
classification system has weaknesses that prompted the above activities. Furthermore, the
current system is primarily designed to denote the perceived primary hazard of a material as
determined by application of rigorous classification criteria Do not under any circum-
stances assume that the hazard indicated by a warning label or placard attached to a
container is the only hazard associated with the material found therein.
8.3 U.S. ENVIRONMENTAL PROTECTION AGENCY CLASSIFICATIONS
The EPA has developed several lists of chemicals and chemical wastes that may be
broadly categorized as"hazardous substances." Besides the water pollutants discussed earlier
in Chapter 6, they include:
• A list of specific hazardous wastes and criteria for designating other
materials as wastes under the Resource Conservation and Recovery Act
(RCRA) of 1976 and subsequent amendments See Title 40, Part 261 of the
Code of Federal Regulations (40 CFR 261) for details.
• A list of hazardous substances developed under Section 311 (b) (2) (A) of
the Clean Water Act of 1977. See 40 CFR 112-114 for details
• Chemicals listed as toxic pollutants under Sections 307(a) and 307(c) of
the Clean Water Act See 40 CFR 116-117 for details.
8-5
-------
FIGURE 8.1
U.S. D.O.T. PLACARDS
The alternate display incorporating the UN/NA 4-digit number appears to the right of the placard
FLAMMABLE GAS
«
FLAMMABLE SOLID
.*
FLAMMABLE SOLID
NON-FLAMMABLE GAS
FLAMMABLE
COMBUSTIBLE
*
CORROSIVE
OXYGEN
CHLORINE
«
EXPLOSIVES
* *
POISON
RADIOACTIVE
EXPLOSIVES
POISON GAS
NOTE No alternate
display Is permitted
for EXPLOSIVES A,
POISON GAS,
and RADIOACTIVE
materials
OXIDIZERS
FLAMMABLE GAS
ORGANIC PEROXIDES
£•
ORGAHIC
.PEROXIDE
PLACARDED EMPTY TANK CARS
NON-FLAMMABLE GAS FLAMMABLE
NOTE
Hazard Class Numbers
1 Explosives
2 Compressed gases
3 Flammable/Combustible
Liquids
4 Flammable solids
51 Oxldlzers
5 2 Organic peroxides
6 Poisons
7 Radioactive materials
8 Corrosives
FLAMMABLE SOLIDS
When required on t«nk cars, portable tanks or cargo tanks, Identification numbers, as specified In §172101 or §172102, shall be dlsplayec
An ldenmicatlon"mimber may not be displayed on a Poison Gas, Radioactive or Explosives placard §172 334(a), but If a tank car,
p^rtabltunk or Mrgotank casing such a commodity requires an Identification number, It must be displayed on an orange panel §172
8-6
-------
• A list of materials deemed to be Extremely Hazardous Substances (EHS)
by virtue of their acute inhalation toxicity in air. Established under Title
IE, section 302(a) of the Superfund Amendments and Reuthorization Act of
1986, this list has been subject to frequent changes. It may be expanded in
the future to include materials with other hazardous characteristics.
• A list of hazardous substances established under the 1980 Comprehensive
Environmental Response, Compensation, and Liability Act (CERCLA),
also known as Superfund The list is comprised of chemicals listed under
RCRA, the Clean Air Act, and/or the Clean Water Act See 40 CFR 302
for details Extremely Hazardous Substances are also to be included in this
list.
• A list of toxic chemicals established under section 313 of SARA Title HI
for emissions reporting See 40 CFR 372 for details
The hazardous substances listed under CERCLA have been assigned reportable
quantities by the EPA. These are the amounts that must be spilled within a specified penod
of time before the party responsible for the spill or discharge is required to report the spill to
federal, state, and local governments They range from one pound for materials considered
to be extremely harmful to the environment (plus some chemicals which are under review
and have not yet been assigned more appropriate reportable quantities) to 5000 Ibs for those
substances considered to pose significant but comparatively moderate environmental hazards.
It is well to recognize that.
• The current EPA CERCLA hazardous substance list mostly includes
substances that were identified as a result of then- long-term environmental
and public health hazards There are many significant hazardous materials
which do not appear in the list
Reportable quantities (RQs) were generally derived from an evaluation of
the long-term health and environmental hazards of the listed chemicals.
RQ values represent a relative ranking of the chemicals vis-a-vis each other
and are not absolute indicators of risk Due to the criteria by which they
were derived, RQs should not be used to rank substances for planning or
emergency response activities involving episodic spills or discharges of
hazardous materials posing acute threats to the public.
Each Extremely Hazardous Substance designated by the EPA has been assigned a
Threshold Planning Quantity (TPQ) which triggers various reporting, community
right-to-know, and emergency planning requirements. Please note that:
8-7
-------
• The list of Extremely Hazardous Substances was prepared quickly after the
Bhopal incident as an attempt to denote those materials which pose a high
acute toxicity hazard to the public when discharged into the environment
The list contains several substances that are highly toxic but lack mobility
under ordinary spill conditions
• Although Threshold Planning Quantities have an important role in defining
regulatory requirements, there is no guarantee that lesser quantities of a
designated EHS will not pose threats to public health and safety under all
accident conditions.
8.4 NATIONAL FIRE PROTECTION ASSOCIATION HAZARD RANKINGS
In an attempt to provide fire service personnel a rapid means of assessing the dangers
of hazardous materials, the National Fire Protection Association (NFPA) has developed a
ranking system that assigns separate values in the range of zero to four to the health,
flammability, and reactivity hazards of individual materials A fourth category for "special"
hazards uses the following symbols among occasional others:
• "W to denote unusual reactivity with water
• "OX" to denote that the material has oxidizing properties
• "COR" to denote that the material is corrosive to living tissue
• The standard radioactivity symbol to denote radioactivity hazards
Table 8.1 defines the rankings specified by the NFPA for health, flammability, and
reactivity. Although the individual rankings are often simply listed by category in NFPA
documents and many chemical company material safety data sheets, they may also be seen
within a diamond-shaped sign with blue, red, yellow, and white squares containing the
respective rankings for health (blue), fire (red), reactivity (yellow) and other (white)
8.5 INTERNATIONAL MARITIME ORGANIZATION CLASSIFICATION
Under the auspices of the United Nations, the International Maritime Organization
(IMO) has developed and continues to refine its International Maritime Dangerous Goods
Code (1MDG) to facilitate and ensure the safety of international shipments of hazardous
materials. The DOT has adopted and/or permits use of IMDG requirements under numerous
circumstances, and it is very common to see references to these requirements in MSDS and
8-8
-------
TABLE 8.1
NFPA HAZARD RANKINGS
oo
Identification of Health Hazard
Color Code: BLUE
Signal
«
8
ft
a
®
Type of Possible Injury
Materials which on very
short exposure could cause
death or maior residual in-
jury even though prompt
medical treatment were
given
Materials which on short ex-
posure could cause serious
temporary or residual injury
even though prompt medical
treatment were given.
Materials which on intense
or continued exposure could
cause temporary incapacita-
tion or possible residual in-
jury unless prompt medical
treatment is given
Materials which on exposure
would cause irritation but
only minor residual injury
even if no treatment is given
Materials which on exposure
under fire conditions would
offer no hazard beyond that
of ordinary combustible ma-
terial
Identification of Flammabihty
Color Code- RED
Susceptibility of Materials to Burning
Signal
«
§
ft
Q
(D
Materials which will rapidly
or completely vaporize at
atmospheric pressure and
normal ambient tempera-
ture, or which are readily
dispersed in air and which
will burn readily.
Liquids and solids that can
be ignited under almost all
ambient temperature condi-
tions.
Materials that must be mod-
erately heated or exposed to
relatively high ambient tem-
peratures before ignition can
occur
Materials that must be pre-
heated before ignition can
occur
Materials that will not burn
Identification of Reactivity
(Stability) Color Code YELLOW
Susceptibility to Release of Energy
Signal
4
8
ft
Q
(D
Materials which in themselves are
readily capable of detonation or of
explosive decomposition or reaction at
normal temperatures and pressures
Materials which m themselves are
capable of detonation or explosive
reaction but require a strong initi-
ating source or which must be heated
under confinement before initiation
or which react explosively with water
Materials which in themselves are
normally unstable and readily under-
go violent chemical change but do
not detonate. Also materials which
may react violently with water or
which may form potentially explosive
mixtures with water
Materials which in themselves are
normally stable, but which can be-
come unstable at elevated tempera-
tures and pressures or which may re-
act with water with some release of
energy but not violently
Materials which in themselves are
normally stable, even under fire ex-
posure conditions, and which are not
reactive with water
-------
other hazardous material publications. Indeed, the DOT has proposed to adopt IMDG
performance oriented packaging requirements in their entirety for implementation in the
United States.
The IMO has categonzed its overall list of hazardous materials into nine major classes,
many of which are further broken down into two or more divisions. Table 8 2 lists and
describes the basic definitions of IMO classes and divisions Detailed definitions, including
more specific breakdowns for explosives, are provided in the text of the IMDG code.
8-10
-------
TABLE 8.2
BASIC IMO MATERIAL CLASSES AND DIVISIONS
Class 1 -- Explosives
Division 1.1 Substances and articles which have a mass explosion hazard Explosive A
Division 1.2 Substances and articles which have a projection hazard but not a mass explosion
hazard. Explosive A or B
Division 1.3 Substances and articles which have a fire hazard and either a minor blast hazard
or a minor projection hazard or both, but not a mass explosion hazard.
Explosive B
Division 1.4 Substances and articles which present no significant hazard. Explosive C
Division 1.5 Very insensitive substances. Blasting Agent
Class 2 -- Gases (compressed, liquelified or dissolved under pressure)
Division 2.1 Flammable gases. Flammable gas
Division 2.2 Nonflammable gases. Nonflammable gas
Division 2 3 Poison gases. Poison A and other poison gas
Class 3 - Flammable liquids
Division 3.1 Low flash point group (liquids with flash points below 0°F) Flammable liquid
Division 3.2 Intermediate flash point group (liquids with flash points of 0°F or above but less
than73°F). Flammable liquid
Division 3.2 High flash point group (liquids with flash points of 73°F or above up to and
including 141°F). Flammable liquid or Combustible liquid
Class 4 - Flammable solids or substances
Division 4.1 Flammable solids Flammable solid
Division 4.2 Substances liable to spontaneous combustion. Flammable solid or, for py-
rophoric liquids, Flammable liquid
Division 4.3 Substances emitting flammable gases when wet. Flammable solid
Class 5 -- Oxidizing substances
Division 5.1 Oxidizing substances or agents. Oxidizer
Division 5.2 Organic peroxides. Organic peroxide
Class 6 -- Poisonous substances and infectious substances
Division 6.1 Poisonous substances. Poison B
Division 6.2 Infectious substances. Etiologic agent
Class 7 — Radioactive substances. Radioactive material
Class 8 — Corrosives. Corrosive material
Class 9 — Miscellaneous dangerous substances. Other regulated material
Note: Corresponding DOT classes are shown in italics following IMO classes and divisions.
8-11
-------
-------
9.0 OVERVIEW OF THE HAZARD ANALYSIS PROCESS
PROBABILITY
ANALYSIS
Likelihood of Accidents
Outcome of Events
HAZARD
IDENTIFICATION
Locations and Routes
Materials and Amounts
Characteristics
RISK ANALYSIS
Combination of
Consequences and
Probabilities
CONSEQUENCE
ANALYSIS
Nature of Hazards
Magnitude of Impacts
PLANNING
FOR
ACCIDENTS
9.1 INTRODUCTION
Chapter 1 to this document reported that recent guidance manuals published by the federal
government have used the term hazard analysis to describe the overall procedure for
evaluating the hazards, consequences, vulnerabilities, probabilities, and nsks associated with
the presence of hazardous materials within any given locality or jurisdiction. This term will
also be used herein for the sake of consistency with earlier publications, although it is
recognized that hazard analysis is often applied in a somewhat different context within
government and industry.
There are four basic steps presented in this guide for the conduct of a hazard analysis, and
a related fifth step that takes advantage of the knowledge gained during the effort to develop a
comprehensive emergency plan for hazardous materials that focuses attention on the known
threats to a community or facility while maintaining sufficient flexibility to deal effectively and
efficiently with unforeseen events These steps include:
• Location, identification, and characterization of potential spill sources and
accident sites in the jurisdiction or locality of concern in a process referred to as
hazard identification. This step essentially concludes with the identification
9-1
-------
and/or postulation of fundamental accident scenarios requiring further considera-
tion and analysis Results from the probability analysis step which follows can
often help in further refining these scenarios.
• Evaluation of the likelihood of individual accident scenarios in a process called
probability analysis. This step permits examination and/or pnontization of
potential accident scenarios in terms of their probability of occurrence.
• Evaluation of the consequences and impacts associated with the occurrence of
postulated accident scenarios in a process that is referred to as consequence
analysis. This step provides an understanding of the nature and outcome of an
accident and permits examination and/or pnontization of scenarios in terms of
their potential impact on people and property.
• Combination of results from the accident probability and consequence analysis
efforts to provide a measure of overall nsk associated with the specific activity
or activities being studied in a process referred to as risk analysis. The effort
permits examination and/or pnontization of scenarios in terms of overall "risk".
• Use of the results of the above activities (which in aggregate provide a planning
basis for emergency preparedness personnel) during actual development and
preparation of an emergency plan.
It is the express purpose of this chapter to introduce and describe these various steps
further and to set the stage for accomplishment of necessary efforts via use of the data,
information, analysis procedures and computational methods presented in subsequent chapters
of this guide.
Note that the various steps of the overall hazard analysis need not be performed m
precisely the order shown for all postulated accident scenarios. Indeed, as descriptions of the
various steps are read, keep in mind that:
• Some users of this guide may wish to employ all steps outlined to one postulated
accident sceneano at a time, starting with the scenano they perceive as posing
the greatest threat to their jurisdiction and then working down their "list".
• Some users may wish to perform one step at a time for all postulated accident
scenarios.
• Some users may wish to ignore the probability analysis step for one or more
postulated accident scenarios if they perceive or determine that the consequences
of an accident would be major or catastrophic and wish to plan for them
9-2
-------
regardless of their likelihood of occurrence. Such decisions are specifically
supported by guidance provided below and in Chapter 13. Although the most
severe yet credible accidents that can be foreseen in any jurisdiction or locality
are most likely to have low probabilities of occurrence, the very fact that
consequences may be catastrophic or major is usually sufficient justification for
consideration of the scenario during the emergency planning process.
• Some users may wish to skip the assessment of accident impacts and conse-
quences for scenarios that are determined to be highly unlikely and are also
known to pose comparatively low threats to the public due to the quantity and/or
characteristics of the materials involved.
• Some users may wish to perform a "quick and dirty" assessment of potential
accident probabilities and/or consequences using readily available information
and assuming "worst case" conditions for unknown data or parameters The
answers obtained could then be used to prioritize more formal analyses of
important accident scenarios
92 STEP 1: HAZARD IDENTIFICATION
Hazard identification involves delineation and specification of those facilities and
transportation modes that handle hazardous materials within the locality or jurisdiction of
concern In other words, it requires that planning personnel determine where and how
hazardous materials are stored, handled, or processed in their locality, how and by what routes
they are transported to and from these facilities, and where and how hazardous materials may
pass through the area on their way to other destinations via rail, highway, marine, or pipeline
transportation routes.
A directly related and important activity involves characterization of each potential spill or
accident site in sufficient detail to formulate potential accident scenarios and to permit
subsequent evaluation of accident probability, likely spill amount, and nature and magnitude of
resulting impacts In other words, once detective work has discovered where hazardous
materials are located, this step involves gathering the data and information necessary to
eventually postulate the circumstances under which accidents may occur and to evaluate the
approximate hazards and risks that these accidents may pose to surrounding populations
Specific guidance and advice pertaining to the conduct of a hazard identification effort
follow in Chapter 10 of this guide.
9-3
-------
93 STEP 2: PROBABILITY ANALYSIS
With all the media attention given to the topic on a national as well as international basis
in recent times, one might easily come to believe that a major disastei involving hazardous
materials is bound to occur within the foreseeable future wherever such materials are handled,
stored, processed, or transported Indeed, a survey of public authorities a few years ago placed
such events at the top of a list of concerns for conceivable emergencies in their respective
jurisdictions Fortunately, however, catastrophic spills or discharges, i e, those that actually
fall or significantly injure more than a few people at a time, are actually rare events in our large
and heavily industrialized nation, although accidents in general involving hazardous materials
are very common The vastly increased attention given to chemical safety in recent times by
industry and government alike should serve to further improve overall safety performance in
the future. Better preparedness to respond to accidents should serve to reduce overall risks to
society by helping to reduce or limit adverse impacts once an accident has occurred
The probability analysis step may be considered optional where community leaders or
facility owners wish to prepare for every conceivable accident regardless of its probability of
occurrence and have the time, manpower, and resources to achieve their goals More often
than not, however, emergency planners will find that time and resources are limited, that other
threats to the community or public needs compete for attention, and that there is value in
conducting a probability analysis Pnontization of chemical related threats in terms of
probability permits attention to these threats in an orderly fashion and reduces the chance that
time and resources will be expended on scenarios of exceedingly low credibility or
significance.
The task of evaluating the potential for a hazardous material emergency in any locality or
jurisdiction involves use of historical accident data in conjunction with local factors (to the
extent possible) to predict the frequency of future accidents, and to some extent, the general
consequences of these events. Prediction of the future, of course, is an inexact science, but
probabilistic accident assessment methods can provide approximate indications of the number
and nature of accidents expected on average in a given locale within a specified penod of time,
and can therefore provide valuable guideposts for decision-making purposes
There are many localities where the total traffic and use of hazardous materials pose a
clear threat to public health and safety and which are generally aware of the need for
comprehensive emergency planning These localities could benefit from a probabilistic
assessment of accident potential which permits the various threats to be ranked and prioritized,
thus ensuring that the most important and serious threats receive the full attention they deserve
and that available resources are wisely allocated (Note- At least one instance is known where
a city purchased a set of expensive chemical protective clothing -- fully encapsulating suits
9-4
-------
resembling space suits ~ and stored them away after allowing its hazmat response team
members to try them on once The suits eventually mildewed and rotted from lack of need and
attention.)
At the opposite end of the spectrum are localities which face relatively few hazardous
material threats and which may be unsure whether the nsk of an accident warrants extensive
expenditures of time and resources for emergency preparedness A probabilistic assessment of
accident potential, coupled with the results of a consequence analysis, can assist these localities
in deciding upon the appropriate level of planning and preparedness. Together with assess-
ments of other natural and man-made threats to the locality, priorities can be set for allocating
time and resources to threats with the highest potential for harming the public. Efforts can be
initiated for sharing response capabilities and resources among neighboring jurisdictions where
the chances of a significant accident in a region encompassing several jurisdictions are
considerable, but the chance that the accident will occur in any specific locale within the region
is comparatively low.
Guidelines and methods for probabilistic assessment of hazardous materials emergencies
are presented in Chapter 11 of this guide
9.4 STEP 3: CONSEQUENCE ANALYSIS
Probabilistic assessment of accident potential can provide a good idea of the likelihood
that a potential accident will actually take place It must be realized, however, that the most
frequent types of spills or discharges have relatively minor consequences, and that more
serious accidents will generally have lower probabilities of occurrence. Thus, a full under-
standing of the risks faced by any specific locale requires knowledge not only of the
probabilities associated with different types of accidents, but also the expected impacts and
consequences of these events
Estimation of potential accident impacts and consequences can be accomplished via a
variety of consequence, vulnerability, and hazard assessment methodologies described in
Chapter 12 of this guide and incorporated within the computer program named ARCHIE that is
an integral part of this document
9.5 STEP 4: RISK ANALYSIS
The nsk analysis step is also somewhat optional in the sense that it relies upon the results
of the accident probability analysis for completion. It entails combination of the probability or
likelihood of an accident occurring with a measure of the predicted consequences of the
accident to provide an overall measure of risk that can be used for threat pnontization and
planning purposes.
9-5
-------
Readers should be aware that the term risk is often misused by society. Being usually
defined as a combined measure of the probability and seventy of potential threats, it is
possible for a threat with low probability of occurrence and relatively high potential seventy to
pose a comparable level of risk to a community (from a planning perspective) as a threat with a
higher probability of occurrence but lower seventy. Thus, the performance of a nsk analysis
permits all threats to be viewed from a perspective that is not biased by consideration of either
probabilities or consequences alone.
Chapter 13 provides guidance on how the accident scenarios evaluated in Steps 2 and 3
may be evaluated in terms of nsk It also contains a discussion of how the nsks associated
with hazardous materials compare with more common threats to life and property The latter
topic is considered important because the hazardous matenals accident problem has several
emotional and political aspects that sometimes tend to distort the truth.
9.6 STEP 5: USE OF HAZARD ANALYSIS RESULTS IN EMERGENCY PLANNING
The scenarios resulting from the overall hazard analysis process will hopefully represent
the full range of significant hazardous material emergencies that have a reasonable likelihood
of occurring in the foreseeable future within any given locality. It remains to consider how
these scenarios and related analysis results may be used to focus an emergency response plan
on credible threats to the locality of concern and to ensure that the emergency plan provides for
efficient, rapid, and comprehensive mitigation of adverse impacts
Chapter 14 of the guide discusses the planning ramifications associated with individual
accident scenarios in some detail and serves as a guide for the use of these scenarios during
emergency planning. Note that each scenario deemed credible and worthy of consideration
gives planning personnel the opportunity to sit back under non-emergency conditions, identify
steps that must be taken to protect the public, and ensure that response personnel will have the
necessary organization, communications systems, equipment, matenals, manpower, sources of
assistance, and training to cope with the situation and minimize casualties, property damage,
and environmental pollution.
9-6
-------
10.0 HAZARD mENTIFICATION GUIDELINES
©xgr
10.1 INTRODUCTION
The purpose of this chapter is to assist planning personnel in identifying and
characterizing potential sources and locations of hazardous material spills within their
jurisdiction. It is primarily directed to local and county governments, but broad application
of its guidance also permits use at the state level and within industry.
The chapter outlines a variety of methods to obtain the desired information. It is left to
individual localities or jurisdictions to select the method or combination of methods best
suited to local conditions
10.2 REASON FOR THE DESIRED INFORMATION
There are three fundamental and interrelated reasons why a town, city, county, or state
government should have knowledge of the identity, location, and characteristics of hazardous
materials and related processes within its boundaries.
1. For hazard assessment purposes: The desired information, together with
identified accident scenarios and the use of consequence analysis procedures
presented in this guide, can provide emergency command personnel with an
-------
indication of the potential nature and magnitude of hazardous material threats
facing a community. This knowledge in turn can facilitate decisions concern-
ing protection of the public and on-scene response personnel in the event of
an actual emergency.
2. For emergency planning purposes: It can be difficult and extremely
inefficient to plan and prepare for every conceivable emergency situation The
desired information, together with the probability analysis and consequence
analysis procedures presented in this guide, permit emergency plans to be
"tailored" to the specific threats facing a community
3. For actual response purposes: Hazardous material spills are often confus-
ing and dangerous situations in their initial stages, especially if responding
emergency personnel do not have a good idea of the nature and quantities of
the substances that may be involved upon arriving at the accident scene. The
hazard identification process permits compilation of a centralized data base
that can be accessed upon first notification of an emergency to determine (or
at least limit) the overall range of possibilities.
10.3 SUGGESTED SCOPE OF THE EFFORT
The guidance that follows may suggest to some readers that the collection and
compilation of the desired data will require a major effort on the part of planning personnel
This will be true to some degree in highly industrialized communities, but the effort can be
made manageable by keeping certain thoughts and concepts in mind.
• Small amounts of hazardous materials (unless they have unusual and
extremely dangerous properties) are generally likely to cause problems in
only a very localized area. Data collection efforts can be greatly minimized
by concentrating efforts on transportation routes and facilities that handle or
store significant quantities of hazardous materials
• Industrial concerns that manufacture or use large amounts of hazardous
materials are likely to employ or have access to technical personnel with
expertise in chemical safety and will be well aware in most cases of their
Labilities for any deaths, injuries, or property damages caused by an accident.
Some facilities, particularly those associated with major corporations, may be
willing to compile the desired data and even perform the analyses described in
Chapters 11,12, and 13 of this guide if asked to do so at the appropriate level
of management These firms have a clear and vested interest m ensuring that
the local community is well prepared to protect the health and property of the
public in the event of an accident. Many of them, especially since the Bhopal
10-2
-------
tragedy and its attendant litigation, have already taken steps to assess and
reduce the nsks that their chemical-related activities may pose to nearby
populations and the continued viability of their business operations in the
event of a major accident.
Major accidents are fairly rare events, particularly when one focuses on any
relatively small part of the nation, as is the case with the 4000 or so local
emergency planning committees (LEPCs) that have been established in
response to federal laws and regulations relating to the Superfund Amend-
ments and Reauthonzanon Act (SARA) of 1986 The detailed and time-con-
suming work can be spread out over some reasonable period of time once
minimum planning requirements mandated by SARA have been fulfilled.
Everyone has a stake in hazardous materials safety. Civic-minded citizens,
business organizations, and individual companies may be willing to volunteer
time and resources to the overall effort. The fact that LEPCs have been
established across the nation has set the stage for, should facilitate, and should
indeed encourage planning efforts that transcend the limitations of mandatory
planning requirements.
Neighboring communities and jurisdictions will need much of the same data
and information, not only on hazardous material transportation traffic, but
also with respect to facilities that may pose a threat across junsdictional
boundaries in the event of an accident Cooperation and integration of
activities on a regional basis may not only reduce the workload for all parties,
but result in cooperative agreements that may increase the effectiveness and
efficiency of emergency response actions during actual emergencies Fire
departments across the nation, for example, have long appreciated the value
of regional mutual aid systems. This concept can be extended in a variety of
ways for response to hazardous material related accidents, thus reducing the
burden on individual jurisdictions.
Many of the efforts and tasks described in this and the following chapters
appear more complex when first looked over than they really are. The work
will go much more smoothly and quickly as experience is gained in applying
suggested methods and interpreting their results
Most importantly, federal laws and regulations require many (though not all)
facilities that utilize hazardous materials to provide LEPC's all information
needed (with certain exemptions related to trade secrets) for preparation of
emergency plans Those facilities that store quantities of Extremely Haz-
ardous Substances (EHSs) in excess of designated Threshold Planning
10-3
-------
Quantities (TPQs) are required to appoint a. facility emergency coordinator to
assist the LEPC in its planning efforts Other facilities may also be required
by SERCs to participate in the planning process under Tide HI of SARA
10.4 NATURE OF DESIRED INFORMATION
Table 10.1 briefly summarizes the types of information generally needed for various
transportation modes and stationary facilities, while subsequent discussions provide further
descriptions of data requirements Both the table and associated discussions are somewhat
general because the specific details needed are not only a function of the properties of the
hazardous materials being handled, but the circumstances under which they are transported,
stored, transferred, or processed, and the degree of detail that emergency planning personnel
wish to include in their overall effort. It is therefore strongly recommended that Chapters 11
and 12 of this guide be carefully studied and that examples be worked out so that data
collection personnel will have (or can be given) specific guidance with respect to the detailed
information desired m any given situation or jurisdiction
Rail Transportation
If the locality of interest has one or more railroad nght-of-ways, it is first necessary to
determine whether these tracks are used for shipments of hazardous materials. Any track
segments used for this purpose should then be characterized in terms of specific location and
length. Special attention should be given to identifying track segments that pass over or
along the side of bodies of water. For subsequent response planning purposes, particularly in
rural areas, some thought should be given to how various portions of the route may be
accessed or approached by emergency response personnel and vehicles Any information
collected on these topics can best be shown on maps of the area, which may also be modified
to highlight population centers and special occupancies such as schools, hospitals, prisons,
and nursing homes ~ not just for railroad accident purposes, but for all credible accident
scenarios.
There are maj'or railroad corridors in the United States that provide passage for a wide
variety of hazardous materials, and there are numerous routes that only have limited traffic to
specific destinations. Some part of this traffic may consist of regularly scheduled shipments
(e.g., weekly or monthly shipments from a particular shipper to a particular receiver), another
part may consist of non-regular but recurrent shipments, and a smaller part may consist of
unique, non-recurrent shipments Ideally, it is desired that planning and emergency response
personnel obtain as complete a picture as possible of the specific hazardous cargos handled
over the course of a recent 6-12 month penod, the types and capacities of their containers, the
general frequency of individual shipments, and the frequency of trains (called "consists" in
the railroad industry) which haul hazardous cargos Such data are best obtained by initiating
direct contact with personnel within the safety department(s) of the railroad(s) that use the
10-4
-------
TABLE 10.1
SPILL SOURCE CHARACTERIZATION FACTORS
Rail Transportation
• Route(s) and associated mileage through locality
• Classifications of track
• Location and layout of railroad yards
• Specific hazardous cargos
Number of cars passing through and length of stay m yards
Types and capacities of containers
Highway Transportation
Route(s) through locality
• Nature and length of roads by segment
• Location and layout of local terminals
• Specific hazardous cargos
Number of trucks passing through and length of stay
Types and capacities of containers
Water Transportation
• Route(s) through locality
• Mileage of route(s)
• Nature and characteristics of waterway(s)
Location and layout of moorings and anchorages
• Specific hazardous cargos
Number of ships or barges passing through and length of stay
Types and capacities of vessels and containers
Pipeline Transportation
• Pipeline route(s)
Mileage of route(s) through locality
• Contents of pipehne(s)
• Pressure and temperature of pipelme(s)
• Flowrate through pipelme(s)
Overall length and diameter of hne(s)
• Characteristics of leak detection and shutdown system(s) (if any)
10-5
-------
TABLE 10.1 (Cont.)
SPILL SOURCE CHARACTERIZATION FACTORS
Chemical and Petroleum Bulk Processing Facilities
• Location and layout of overall facility
• Location, type, dimensions, capacity, venting systems, contents, pressure, and tempera-
ture of chemical reactors, storage tanks, holding tanks, and other vessels
• Route, length, diameter, flowrate, pressure, temperature, and contents of major
intraplant pipelines, together with information on leak detection and shutdown systems
• Location and nature of bulk cargo loading and unloading facilities and frequency of
transfer operations
• Location, size, and layout of secondary containment systems such as sumps, trenches,
dikes, or barriers around potential spill locations
• Location, layout, and destination of sewer and drainage systems
Chemical and Petroleum Bulk Storage Facilities
• Location and layout of overall facility.
• Location, type, dimensions, capacity, pressure, temperature, venting systems, and
contents of bulk storage tanks.
• Route, length, diameter, flowrate, pressure, temperature, contents, and frequency of use
of major intrafacihty pipelines together with information on leak detection and
shutdown systems.
• Location and nature of bulk cargo loading and unloading facilities and frequency of
transfer operations.
• Location, size, and layout of secondary containment systems, as above
• Location, layout, and destination of sewer and drainage systems.
Pesticide and Other Packaged Chemical Warehouses
• Location and layout of overall facility.
• List of products stored in f acility together with data on storage locations
• Fire protection systems in the facility
• Location, layout, and destination of sewer and drainage systems in area.
• Location, size, nature, and layout for secondary containment of spilled chemicals or
contaminated water used for flrefighting
10-6
-------
track(s) in question. Although these individuals may refer you elsewhere within their
organizations for the desired data, contact with them can also be beneficial because they will
have knowledge of the plans and preparations undertaken by the railroad to respond to
emergencies involving hazardous commodities along their routes -- information which will
be highly useful during the community emergency planning process. (Note: Railroads are
common carriers, as are many trucking firms and marine shipping operations. Although
such earners do not own the cargos they transport, and very often do not own the vehicles in
which the cargo is placed, they are legally and financially responsible for the damages
resulting from any accidents that occur while a hazardous commodity is in their possession.
As common earners, they in most cases have little choice but to transport any commodity
placed in their custody in accordance with current federal regulations.)
If the locality includes a railroad terminal or yard that transfers cars with hazardous
cargos between trains heading to different destinations and/or which stores them for a time
(which could be days or more) before they move on, a map should be obtained of the specific
location and layout of the facility. In addition, if hazardous material cars are in any way
segregated or sorted into special holding areas, these areas should be identified. As in the
case of moving traffic, as much information as possible should be obtained as to the
identities, quantities, frequency, and length of residence of individual cargos over a
representative period of time. Employees at the facility are the best initial source of
information Be sure to ask if they have prepared an emergency plan for the facility and if
they are willing to integrate and/or coordinate this plan with the general community planning
effort.
Highway Transportation
Much of the information desired for railroad earners of hazardous matenals is also
desired for over-the-road earners Specific needs include identification of major traffic
corridors; specification of the location, length, and nature of roads; characterization of
hazardous cargos, shipment frequencies, container types, and container capacities, and
characterization (as in the case of railroads) of any local terminals or other gathering areas
for hazardous material transport vehicles such as truck stops, weigh stations, motels, and so
forth. It may also be beneficial to compile data on any travel and route restrictions in effect
in the region.
Water Transportation
Those localities on the coastline of the United States, bordering inland waterways, or
home to a port or harbor must be concerned with haulage of hazardous matenals by ship or
barge. Although relating to the marine environment, the desired information is similar to that
discussed above for railroad and highway transport.
10-7
-------
Pipeline Transportation
Most major cross-country pipelines in the United States convey natural gas, crude oil,
LPG, refined petroleum products or anhydrous ammonia, but one may occasionally
encounter more exotic commodities being transferred between specific sites (e.g., between
the manufacturer and a major user of a particular chemical) Besides the specific route of a
pipeline and its length through the jurisdiction of concern, it is desirable to determine
pipeline contents, pressure, temperature, typical and maximum flowrate, diameter, and
overall length, as well as characteristics of any leak detection and shutdown systems. The
effort should probably include high pressure natural gas transmission lines and larger
distribution lines, but not the smaller low pressure pipelines serving individual buildings and
neighborhoods.
Bulk Chemical and Petroleum Processing Facilities
This category includes a variety of chemical manufacturing plants, oil refineries, and
facilities which use large quantities (i.e, bulk quantities) of hazardous materials during the
manufacture of their products In order to make the information gathering task more
manageable at large and complex facilities, it may be necessary to screen chemical handling
areas and focus on those that utilize the largest quantities of the most hazardous materials,
keeping in mind the difference between the toxicity of a substance and the toxic hazard
presented to the public. Plant personnel, if cooperative, can be of invaluable assistance in
this task. Screening can also be facilitated by asking plant personnel for a list of hazardous
materials used at the facility together with typical quantities on hand.
Of key importance is the need to obtain a plot plan of the facility showing the location
of hazardous material stores (tanks, loading racks, pipelines, etc.) as well as the identity and
amount of chemicals present This information is useful for actual emergency response as
well as planning purposes. Aerial photographs can also be valuable.
It is next prudent to focus on any storage tanks of chemicals or large containers used
for mixing or reacting chemicals. Desired information includes type and location of the tank
or container, working capacity, dimensions, maximum potential pressure and temperature of
contents, identity of contents, the discharge orifice size of any emergency pressure relief
vents, and any systems installed to capture, recover, or destroy flammable and/or toxic gases
that may be released under emergency conditions. While looking at such tanks, it is also
important to determine if the contents have the ability to undergo any type of runaway
exothermic polymerization or decomposition reaction, or if they are subject to any other
hazardous reaction in the event of equipment failure, human operator error, or inadvertent
contamination by materials available in the general area of the tank or container (particularly
10-8
-------
those that are somehow linked to the tank by a pipeline system). Make note also of the
diameters of pipelines and various pipes that exit each tank and which could discharge the
contents of the tank or container in the event of a break or rupture.
While on the subject of intraplant pipelines, note that such facilities may have literally
miles of relatively small pipes linking items of equipment together and possibly a limited
number of intraplant lines of large capacity. It is well to gather the list of information under
Pipeline Transportation above for the larger lines conveying hazardous materials.
Facilities that handle large volumes of chemicals are also likely to ship and/or receive
hazardous materials by rail, highway or water transportation modes, thus necessitating cargo
loading and unloading facilities. Details of interest at these sites include identity, frequency,
and volume of individual shipments, diameter and type of loading/unloading hoses or arms,
normal pumping rates of cargos through hoses or arms, time required to observe any tank
overflows and to shutdown pumps, type and capacity of transport vehicles, and the number,
contents, and duration of stay of railcars, trucks, or marine vessels serving the facility.
Many facilities have installed secondary containment systems around storage tanks,
process areas, and loading/unloading areas to collect and contain any spilled materials,
typically in the form of dikes, curbs, or other barriers surrounding items of equipment or pits
or sumps to which spilled cargo will flow and collect Since the total rate at which vapors or
gases will evolve from a pool of boiling or evaporating liquid is a function of the surface area
of the pool, and since the size of a liquid pool fire is also a function of pool dimensions, the
dimensions of diked areas or sumps provide highly useful information For response
planning purposes, information is desirable on plant fire protection systems and emergency
spill response capabilities.
Finally, it is also desirable to determine whether spilled materials have the opportunity
to enter underground sewer or drain pipes and where these pipes lead. Flammable liquids in
pipes that are not full and exposed to a source of air can evolve vapors that might explode
violently if ignited. Either flammable or toxic substances might pose problems at the end of
the pipe, be it a nver, lake, or water treatment facility.
Chemical and Petroleum Bulk Storage Facilities
There are various bulk storage facilities across the United States that receive bulk
quantities of chemical and petroleum products, including liquefied energy gases such as LPG
and LNG, by one or more modes of transportation, store them for a time, and then load them
onto other transport vehicles for distribution to buyers and users of the products.
10-9
-------
The information desired about such facilities is the same as that needed for the portions
of bulk chemical and petroleum processing facilities that store and transfer products. Refer
to the previous section for a discussion of specific items.
Pesticide and Other Packaged Chemical Warehouses or Distribution Centers
Numerous types of common businesses store hazardous materials in bottles, drums,
boxes, cylinders, and other packaging materials. Although the amount in any given package
might be relatively small, the facility may store a large number of such packages in some sort
of warehouse facility or even on the shelves of a store of some kind. Some materials may be
in small containers simply because buyers do not need or want larger quantities at any given
time. Others may be hi such packages, particularly at laboratory supply companies, because
DOT regulations prohibit transportation of bulk quantities due to then: special hazards
As in prior cases, it is desired to obtain information on the location and layout of the
overall facility, the products typically stoied therein, their usual storage locations, sewer and
drainage systems in the area, secondary containment systems, and fire protection systems.
The latter two topics are particularly important for this category, because one of the more
worrisome scenarios involves fire. Serious environmental impacts may occur if large
amounts of water applied to burning chemicals cause contaminated runoff into sewers or
drains leading to bodies of water or treatment plants Indeed, some fire departments have
prefire plans, particularly for warehouses storing toxic pesticides, that call for allowing the
facility to burn while only protecting surrounding buildings with water until all chemicals are
consumed, thus avoiding a water contamination problem. Although the building may be a
total loss, and populations subject to smoke exposure may require evacuation or other
protective action, savings may actually be realized because of the expenditures that would
otherwise be necessary to decontaminate land and water bodies polluted by contaminated
runoff.
Miscellaneous Facilities
Besides the large number of facilities and transportation modes that are commonly
associated with the chemical and petroleum industries, there are many other common types
of businesses and facilities that are apt to use or store hazardous materials and which should
not be overlooked Some possibilities are listed in Table 10 2 but are only a sampling of the
many types of facilities likely to store and use hazardous materials in some significant
quantity.
10-10
-------
TABLE 10.2
MISCELLANEOUS POTENTIAL SPILL SOURCES
Airport and marine fuel depots - gasolines and fuel oils
Breweries and distilleries - alcohols
Compressed gas suppliers - medical and industrial gases
Construction firms and sites - explosives, compressed gases, fuels
Dry cleaners - cleaning solvents, perchloroethylene
Electronic circuit makers - acids
Embalming supply houses - formaldehyde
Farm/garden supply shops - pesticides, fertilizers, herbicides
Fireworks manufacturers - explosives, pyrotechnics
Food stores or warehouses - ammonia (in refrigeration systems), combustible dusts
Foundries - resins, other chemicals
Fuel oil companies - fuel oils
Furniture snipping operations - solvents
Gasoline stations - gasoline
Gun and ammo shops - ammunition, explosives
Hazardous waste disposal facilities - virtually anything
Hospitals - compressed gases, medicines, radioactive materials, ebologic agents
Laboratories, chemical and biological - various chemicals, etiologic agents
Lawn fertilizer companies - pesticides, herbicides, fertilizers
Leather tanners - various chemicals
LP-gas or propane suppliers - liquefied flammable gases
Paint, varnish, and lacquer makers and wholesalers - resins, solvents, chemical pigments and additives
Pest control companies - pesticides, poisons
Plastic and rubber makers - solvents, additives, bulk chemicals
Plating shops - acids, cyanides
Pulp and paper mills - bleaches, caustics, acids, sulfur compounds, and others
School and university chemical laboratories - various chemicals
Swimming pools (public) - liquefied chlorine
Swimming pool supply houses - oxidizers (calcium hypochlonte), hydrochloric acid, algaecides
Steel mills - acids, degreasers
Textile and fiber manufacturers - solvents, dyes, resins, various other bulk chemicals
Water treatment facilities - liquefied chlorine, acids
Welding shops - compressed gases
Welding supply shops - compressed gases
10-11
-------
10.5 AVAILABLE METHODOLOGIES TO COMPILE DESIRED INFORMATION
Previous sections have discussed the primary reasons for identifying and characterizing
potential spill sources and accident sites, and briefly described the information and data
desired for hazard identification and analysis purposes. It is now time to briefly review some
of the methods available for fulfilling these information needs. In reviewing this material,
keep in mind that major potential spill sources outside the the community of interest may
also have the capability to impact residents and their property in the event of an accident It
may not be enough to only study facilities that he within the precise boundaries of the
community or jurisdiction of concern
Enforcement ofRight-to-Know Laws
Right-to-Know laws or regulations, be they of federal, state or local origin, typically
require that manufacturers and users of specified hazardous materials provide material safety
data sheets (MSDS) or lists of available MSDS for the substances handled on their respective
sites to employees, community leaders, fire departments, state emergency planning groups,
and/or members of the general public. The various laws and regulations enacted or
promulgated over the years have differed in several important aspects, but all, in some
fashion or another, have had the potential to facilitate identification of facilities which handle
or otherwise utilize specified hazardous materials.
Although Right-to-Know laws were enacted in more than 25 states in recent years,
recently revised federal laws and regulations have essentially preempted most of these
legislative initiatives. The new laws and regulations are very comprehensive and have the
net objective of ensuring that State Emergency Response Commissions (SERCs) and LEPCs
will be automatically informed of the piesence of most hazardous materials present at
facilities within their respective jurisdictions. Indeed, enforcement of right-to-know laws
and regulations is the most direct and efficient method available to LEPCs for the
identification of facilities that manufacture, store, process, or otherwise use hazardous
materials that may pose a threat to community health and safety. In most cases,
enforcement may require little more than informing these facilities, either individually or via
a general public relations campaign, that they are subject to these laws and regulations
Although progress is being made in this area, there remain many facilities and businesses,
particularly those that do not consider themselves part of the chemical industry, who have
been slow to realize or recognize that they are fully subject to the mandates of these laws and
regulations regardless of the nature of then: operations.
The specific reporting requirements originally mandated by SARA are summarized
within Appendix A of the Hazardous Materials Emergency Planning Guide (NRT-1) cited
in Chapter 1 of this document. Stated quite briefly, SARA requires that facilities storing or
using EHSs in excess of TPQs must notify the local SERC and LEPC, appoint a facility
10-12
-------
emergency coordinator to assist the LEPC in emergency planning and to provide any
additional information and data required during the planning process In addition, any
facility subject to the the Hazard Communication Standard (29 GFR 1910 1200) of the
Occupational Safety and Health Administration (OSHA) must submit a list of the hazardous
chemicals on its site or a set of material safety data sheets for these materials to the state
emergency response commission (SERC), local emergency planning committee (LEPC), and
local fire department These organizations are also to be provided annual reports of
hazardous material mventones grouped by hazard category Because SARA Title III makes
planning for EHS's a mandatory effort, hazard identification should begin with these
materials.
A very significant and somewhat recent development (late 1988) is that OSHA
succeeded, after a battle in the courts, to expand coverage of its hazard communication
standard from a very specific and somewhat limited set of industries to all employers except
those in the construction industry Thus, many loopholes (though not all) that would
otherwise have complicated the hazard identification efforts of LEPCs have now been closed
by the federal government
The various laws and regulations discussed above are being frequently modified and/or
expanded in coverage in a concerted attempt to further facilitate the work of local emergency
planning personnel. Current information on federal initiatives under the Resource Conserva-
tion and Recovery Act and Superfund law may be obtained by calling the RCRA/Superfund
Hotline at 800-424-9346 or 202-383-3000 Current information on the specific regulations
prompted by Title IH of SARA can be obtained by calling the Emergency Planning and
Community Right-to-Know Hotline at 800-535-0202 or 202-479-2449 Both hotlines have
been established by the federal government and are operational from 8 30 am to 7 30 pm
EST during the normal work week.
NOTE WELL: Although the above laws and regulations will greatly help community
emergency planning personnel in identifying fixed-site facilities that handle hazardous
materials, they are not necessarily all inclusive and encompassing There will still be many
cases in which hazardous materials are handled at a facility but insufficient data will be made
automatically available to LEPCs to permit the performance of a comprehensive hazard
analysis. Additionally, the fact that reporting requirements do not apply to transportation
modes conveying hazardous materials through individual jurisdictions is highly significant
and important, as is the fact that many fities and businesses are not yet aware of their
specific responsibilities Consequently, comprehensive planning for hazardous materials
emergencies, although not fully mandated by law, requires a concerted effort on the
part of LEPCs to identify and characterize potential threats that have in one way or
another escaped mandatory reporting and planning assistance requirements.
10-13
-------
Use of Fire Department and Building Inspection Records
Over the years, local fire departments may have accumulated substantial data on the
businesses and facilities within their jurisdiction as a part of fire hazard surveys, inspections
associated with regulatory or insurance requirements, response to accidents, and preplanning
for fire emergencies. It follows that fire department records and personnel can be a key
source of desired information hi many communities and counties Similarly, building
inspection departments of local governmental entities may have useful records and knowl-
edgeable personnel And last but not least, note that local police departments will have
considerable general knowledge of the businesses that operate in their respective jurisdic-
tions. Law enforcement personnel patrol the streets on a 24-hour basis and will often have
first hand knowledge of many of the potentially hazardous activities of concern
Use of Industry Questionnaires
A reasonably detailed questionnaire mailed to all businesses which may handle
hazardous materials, particularly if accompanied by a letter from the mayor, town or city
council, or local fire chief, can provide valuable information on a significant fraction of the
facilities contacted. It is a good idea to call first to determine to whom the questionnaire
should be directed, and also, to determine who can be called for follow-up questions.
Alternatively, a self-addressed, stamped postcard can be inserted in the package with a
request that the person given responsibility for completing the document list his or her name
and telephone number on the card and drop it in the mail A news release to local media
about the effort can alert the business community as to the arrival of the questionnaire and
alert the public about the planning effort being undertaken. In all cases, be sure to stress the
fact that the information is solely desired for emergency response planning purposes. As
discussed later, be sensitive to the possible need to maintain the confidentiality of certain
data.
Meetings with Business Organizations and Trade Groups
Many businesses throughout the country are members of local Chambers of Commerce
or mutual aid groups (i e., groups of companies in the same industry that have agreed to help
each other as best they can in time of emergency). Presentation of community emergency
response planning information needs during a general meeting of such a group has the
potential to obtain publicity for the effort, to obtain assistance and cooperation from the local
business community, and most importantly, to obtain formal endorsement of the organization
for the effort, thus encouraging individual members to cooperate fully with public authorities.
10-14
-------
Meetings with Individual Business Personnel
Where the effort is within reason, planning personnel assigned this task may choose to
meet with key personnel within the individual companies or organizations handling
hazardous materials, explain the benefits of cooperation to both the community and the
business, and request a tour of facilities. Necessary details of operations can be obtained
during these meetings and tours, or questionnaire forms may be left behind for later
completion.
It is well to keep in mind during such meetings or other contacts that many companies
have undertaken intensive efforts to determine the hazards and nsks faced by the community
and themselves due to the storage and use of hazardous materials, especially hi the aftermath
of the Bhopal incident. Requests for the results of such analyses might lead to the receipt of
much desired information. Given that the Chemical Manufacturers Association is actively
encouraging such efforts among the entire chemical industry under its Community Aware-
ness and Emergency Response (CAER) Program (see Chapter 1), a request may even provide
impetus for initiating such work, which could ultimately save community planning personnel
considerable effort.
It is also desirable to ask if the facility has a contingency plan for dealing with on-site
emergencies and whether any attempt has been made to coordinate and integrate the plan
with community efforts. This can prompt some thought on preparing a facility plan where
none exists, lead to obtaining a copy of the plan (which is bound to contain useful
information), and/or initiate a useful and continuing dialog between company and public
emergency preparedness personnel.
As noted earlier, it is important during such meetings to stress that the ultimate
objective of the community is to ensure it is prepared to protect the public during potential
hazardous material emergencies and to lend appropriate assistance to the responsible party
(i.e., the company or firm that owns and/or has custody of the materials in question) in
mitigating damages resulting from an accident. It is also important, however, to be sensitive
to the fact that the success and commercial viability of some businesses may depend on
proprietary technology or processes that cannot fully be protected with patents or copyrights
Indeed, one or more products of any chemical-related business may be based on "secret
recipes" that would hurt the company if disclosed to competitors. Do not be surprised,
therefore, if there is reluctance at times to discuss certain details of company operations
The right-to-know laws and regulations discussed above have specific provisions
relating to claims of trade secrets by facilities and these provisions should be followed when
applicable. When not applicable, there are essentially two methods to approach the problem
when there is an acknowledged or known threat to the community and issues of confidentiali-
ty that hamper planning efforts The first involves a formal agreement between the company
10-15
-------
and the community to use any information provided solely for emergency planning purposes
and not to disclose the information to any third parties. This places a substantial legal burden
on the community and requires active management of sensitive data, but potentially serves
the needs of both parties The other approach necessitates that the company itself undertake
the analysis procedures outlined in Chapters 11 and 12 and provide planning personnel with
only the final results and a promise to disclose the identity of any discharged or spilled
materials immediately in the event of an accident. The effort may be facilitated on the part of
the company by the employment of independent consultants or contractors.
Queries of Rail, Marine, and Pipeline Transportation Companies
One of the more difficult tasks in some localities will involve compilation of sufficient
data on transportation of hazardous commodities in the area, particularly if these cargos do
not have a specific destination in the locality that can be identified and queried but are simply
passing through The survey methods to be considered vary with the mode of transportation
being utilized.
As note above, the best source of railroad traffic data is the railroad company that
owns and operates any specific track segment of interest. Many will have computerized
records of train movementss and the cargos earned. Others might be willing to compile the
desired data over a specific penod of time to assist the data collection effort Companies that
receive or ship hazardous materials and which have facilities in the area can provide data on
the portion and nature of the traffic for which they are responsible if cooperative. They can
also act to inform emergency planners of the specific transportation companies that they
employ for potentially hazardous cargo movement.
Pipeline routes, particularly those conveying hazardous commodities, are often clearly
marked and known to public authorities, particularly those who may have issued a permit for
the route at some time in the past Substantial information may be readily available from
"digsafe" program offices in many areas of the country that maintain records of buried
pipelines and cables. Local utilities will know of such programs, as will construction
companies, and a hotline number for the local digsafe program is likely to be prominently
displayed in the local telephone directory Contact with the owner of any pipeline is likely to
provide current operating conditions, specific route of the pipeline, and any emergency
response preparations that have been undertaken in anticipation of a future accident
Sources of information in ports and harbors include the port or harbor master, the
companies that offload, ship, or receive hazardous cargos, the public commission or council
(if any) that operates or has regulatory or oversight responsibility for the waterfront, the
marine transportation companies that operate in the area, and any local or regional offices of
10-16
-------
the U.S. Coast Guard or U.S Army Corps of Engineers. The best approach is to locate a
person who knows how the port or harbor operates and how records are kept, and ask his or
her advice on how best to proceed.
Truck Traffic Surveys
By their very nature, trains, pipelines, and marine vessels follow routes that are fixed in
location, readily identifiable, and utilized by a comparatively small number of transportation
companies. This is not true in the case of truck traffic which may be found on a wide variety
of roads and highways and potentially involve a large number of different carriers. The task
of characterizing truck traffic and its routes will therefore be substantially more time-con-
suming in many jurisdictions but can be accomplished to a large extent using the following
methods and procedures.
Almost all jurisdictions are served by firms that provide fuels such as gasoline, fuel oil,
and LP-gas (LPG) or propane to industry and the public These should never be overlooked
Trucks conveying gasoline are known to be involved in more serious accidents than those
transporting any other hazardous material in the United States due to the flammable and
explosive properties of the substance and its extremely widespread use and distribution.
LP-gas and liquefied propane, which are also used in large quantities in most parts of the
country, are highly flammable compressed liquefied gases that may fuel pool fires, flame
jets, BLEVEs, vapor cloud deflagrations, and confined and unconfined vapor cloud
explosions if discharged to the external environment Fortunately, the vast majority of firms
that receive, sell, and deliver these commodities will be readily identifiable. Those
companies that serve commercial accounts or the general public usually advertise frequently
to gain new customers and will be easily found in the "yellow paces'* of local or regional
telephone companies, as will product wholesalers or distribution companies. Discussions
with the operators of a sample of local gasoline stations can also be helpful.
The shippers and receivers of other hazardous materials in the locality of concern are
one good source of information about the nature and frequency of over-the-road shipments
Routing and additional shipment information can be obtained from the trucking firms that
deliver or pick-up cargos
Most large tracking companies have established terminals at various locations across
the country. Although these terminals may be located outside the locale of concern,
company management, safety or dispatch personnel may be able to provide substantial
information on the cargos routed through the subject jurisdiction They may also have a
good idea of the operations of other earners that function in the area, since keeping track of
the activities and operations of competitors is often good business practice Similarly, smaller
companies in the region, particularly those that specialize in carnage of hazardous materials,
can be a useful source of information.
10-17
-------
Finally, when all else fails to provide a reasonably complete overview of hazardous
materials traffic by highway, there is always the possibility of conducting a roadside traffic
survey. One form of survey involves stationing observers along highways at key locations of
interest, such as major intersections or entrances into the jurisdiction for a period of time and
making counts of traffic. The placards and other signs on individual vehicles (see Chapter 8
for descriptions) will provide a general if not specific indication of the nature of any
hazardous cargos being earned The names of transportation firms on vehicles can provide
leads to sources of more detailed information when needed
The task can be somewhat facilitated for major highway traffic if there is a truck
weighing station on the route and observers are positioned at these locations Similarly,
survey activities may be coordinated at times with roadside safety inspections conducted by
the highway patrol, state police and/or department of motor vehicles Police forces may also
be asked to make note of hazardous materials traffic at all times during routine duties on a
periodic basis, as may collection personnel at toll booths.
Although much of the information of interest for planning purposes has been provided
above, it is well to note that the U.S. Department of Transportation and the U S. Department
of Energy jointly sponsored a study to identify and characterize methods by which
information may be obtained on hazardous material shipments at the local level. The
resulting report is:
• Overcast, T.D, and Dively, D.D., Options for Gathering Information
About Shipments of Radioactive and Other Hazardous Materials, DOT
feeport DOT/RSPA/DHM-61-86-2, 1987, available from the National
Technical Information Service, Springfield, VA 22161.
Use of Permit Records
Companies that handle hazardous materials typically require a variety of permits and
licenses to build their facilities, to handle or store flammable materials, chemicals or wastes,
and to discharge pollutants into the air or water. These permits and licenses are issued by a
variety of local, state, and federal authorities and may provide a reasonably efficient means to
identify these facilities in some jurisdictions. Organizations with possibly useful records
include fire and police departments, building inspection agencies, zoning boards, public or
occupational health and safety departments, state and federal environmental protection
agencies, water and sewer commissions, the U.S Army Corps of Engineers, and the U S
Coast Guard, among others.
10-18
-------
Use of the "Yellow Pages"
The local telephone company "yellow pages" directory can be of major assistance in
identifying all types of firms that potentially store, handle, or transport significant quantities
of hazardous materials. Better yet, it can provide their addresses and telephone numbers
Access to Detailed Chemical Property Data and Hazard Information
As readers will realize, industry is generally required to provide emergency planning
personnel with material safety data sheets (MSDS) or information similar to that found on
such sheets It is important to note that this information and data, although of considerable
value in many respects, is not always adequate for a full and complete understanding of the
hazards and properties of individual hazardous materials. Indeed, it is not uncommon to
review several MSDS for the same material, each prepared by a different manufacturer, and
to find numerous subtle and sometimes major inconsistencies in then- contents
Quite frankly, MSDS with different origins and authors often vary greatly in both
accuracy and completeness, although the situation has improved vastly in recent years Even
the best available MSDS, however, provide but a simplified overview of material hazards,
appropriate first aid measures, suggested emergency response actions, and so forth. It is
often best, therefore, to use MSDS as a screening tool to identify those materials that clearly
pose the greatest hazards to a jurisdiction, and then to make an additional effort to compile
more complete information on these substances Realize also, however, that some manufac-
turers have a tendency to exaggerate hazards of then- products to avoid the possibility that
they will be found negligent in warning customers of possible dangers in the event of an
accident Conversely, though this is not nearly as common as it once was, and may simply
have been due to a lack of knowledge or expertise, some firms have been known to downplay
the hazards of their products in the past
There are essentially three available methods to obtain more detailed information on a
specific hazardous material. The first involves access and study of the numerous hazardous
material data bases, handbooks, and guides available in the marketplace The best of these
contain considerable useful and accurate data on the common and sometimes rare chemical
products and fuels used in industry. The worst, however, can be a considerable waste of
funds, so it is wise to purchase these documents or computer programs with care. (Note.
Some computerized data bases, in particular, have absurdly high prices given the nature and
quality of the data they provide In many cases, the authors have simply copied data that is
available in hardcopy documents selling for a tiny fraction of the computer program price )
The name and addresses of chemical manufacturers and distributors, together with
indexed lists of then- products, can be found in several chemical buying guides. Potential
sources of these guides include major public and university libraries, the purchasing
10-19
-------
departments of companies that buy chemicals, and even individual chemists and chemical
engineers in the community. The second approach involves identification of the manufac-
turers of materials of special concern and the mailing of requests for detailed product
information bulletins and safety handling guides Most large corporations will honor such
requests at no cost.
The third approach, which is best suited to identifying a source of detailed information
for an unusual hazardous material, involves calling the Chemical Referral Center in
Washington, D.C. This Center was established by the Chemical Manufacturers Association
(CMA) in late 1985 to assist callers in contacting sources of information on over 250,000
chemical products and basic chemicals on a non-emergency basis. The center operates from
8 a.m to 9 p.m. (EST) Monday through Fnday and may be reached toll-free from the
continental United States and Hawaii at 800-CMA-8200. Callers in the District of Columbia
and Alaska may telephone 202-887-1315 on a collect basis.
10.6 SOURCES OF ADDITIONAL HAZARD IDENTIFICATION GUIDANCE
The U.S. Department of Transportation and the Federal Emergency Management
Agency have sponsored numerous demonstration projects associated with hazardous materi-
als safety and emergency response planning. The experiences of the local, county, and state
authorities involved in these projects, as documented in the reports they have prepared on
their activities, can provide additional ideas and insights on how best to conduct a hazard
identification survey in any given locale. Most of these reports are listed and referenced in
Appendix E of the Hazardous Materials Emergency Planning Guide (NRT-1).
10.7 FORMULATION OF CREDIBLE ACCIDENT SCENARIOS FOR PLANNING
PURPOSES
As noted earlier, it is not sufficient to determine the locations and characteristics of the
sites and hazardous materials that may become involved in a future accident; it is also
necessary to gain an understanding of the potential nature and outcome of potential accidents
for comprehensive planning purposes. An important step in this process is the formulation or
postulation of credible accident scenarios based on the information obtained during hazard
identification activities. These scenarios can then be evaluated with respect to specific
probabilities of occurrence, consequences, and/or nsks to the community via the procedures
respectively described in Chapters 11,12, and 13 of this guide.
Chapter 11, in particular, describes, discusses, and enables probability analysis for
credible accidents at fixed site facilities and in transportation. Although the probability
analysis step itself is considered optional, planning personnel should nevertheless refer to
Chapter 11 for assistance in the definition of accident scenarios in then- respective
jurisdictions based upon results of the hazard identification procedures They may choose to
10-20
-------
select several scenarios for each hazardous activity (ranging from low to high seventy) for
further analysis, or to simplify matters by only selecting worst case scenarios (i e., those
posing the greatest threats to the community.)
10.8 ORGANIZATION OF THE DATA
As noted previously, the information compiled during the hazard identification and
characterization task is not only useful for planning and hazard analysis purposes, but can
also be valuable during actual emergency response if readily accessible in a centralized data
base. It is therefore a good idea to organize the information and data in a filing system of
your choosing, and/or to enter the most important facts into a data management system
installed on a personal computer. If placed in a location that is manned on a 24-hour basis,
such as police or fire department dispatch offices or a designated emergency operations
center, response personnel can be briefed on potential hazards by radio while en route to an
accident site and be given answers to specific questions that may arise upon their arrival at
the scene.
10-21
-------
-------
11.0 PROBABILITY ANALYSIS PROCEDURES
11.1 INTRODUCTION
The transportation and use of hazardous materials poses threats that are of concern to
society, but which are not always fully understood in terms of their likelihood of occurrence
or viewed in perspective with regard to their relation to other threats The purpose of this
chapter is to provide emergency planning personnel with the basic information, data, and
procedures necessary to refine and evaluate individual hazardous material accident scenarios
in terms of their annual probability or frequency of occurrence on at least an order of
magnitude basis
As noted earlier in Chapter 9, this probability analysis step may be considered optional
where community leaders (or individual facility operators) wish to prepare for every
conceivable accident regardless of its likelihood of occurrence and have the time, manpow-
er, and resources to achieve their goals. More often than not, however, these individuals will
find that time and resources are limited and that other threats or needs compete for attention
Pnontization of potential accidents in terms of annual probability as well as seventy will
permit attention to these threats (and it is eventually hoped all threats to public health and
safety) in a logical and orderly fashion and thereby reduce the chance that time and resources
will be expended on emergency scenarios of exceedingly low credibility or significance
Note, however, that assessment of hazardous material accidents or any other threats on a
probabilmsttc basis does not guarantee that all such hazards will be identified or that less
likely events will not occur
The chapter addresses seven primary activities associated with hazardous materials,
each with the potential to result in public emergencies These include
• Bulk transportation by truck
• Bulk transportation by rail
• Bulk transportation by barge or other marine vessel
• Transportation by pipeline
• Bulk storage, processing, or handling at fixed facilities
• Transportation of packages
• Transportation by air
The overall approach presented in this chapter involves use of average national
accident rates determined from historical records and relevant exposure data While such
rates may overestimate or underestimate average annual accident probabilities for any
specific facility or transportation activity, they are not expected to be vastly in error in
aggregate for any jurisdiction or facility The ultimate goal, after all, is not exact estimation
of accident probabilities, but their approximation at a level of accuracy sufficient for
-------
emergency planning purposes. Appendix F to this guide presents the basis and rationale for
the recommendations and procedures that follow, and additionally, discusses the use of local
and other data and information (where available) to further refine estimates where this may
be desireable.
The actual computation of the annual accident probability associated with any specific
activity involves using the frequency with which such accidents are known to occur in
combination with a measure of the "exposure" of the community or other jurisdiction to the
potentially hazardous activity For most transportation modes, for example, historical
accident rates are presented in terms of the number of accidents expected per mile of travel
and exposures are expressed in terms of the number of trips made per year and the mileage of
routes within the locality. Simple multiplication of these values provides the expected
number of accidents per year involving the activity being considered Further multiplication
of this result by such factors as the fraction of accidents that result in loss of cargo to the
environment and the fraction of accidents that result in a prespecified amount of cargo loss
permits greater specificity in predictions Worksheets for each activity facilitate computa-
tions, and are provided in lieu of tabulated or generalized categorizations of accident
probabilities to provide planning personnel with the option of using local data for greater
accuracy. Planning personnel may use the predicted accident frequencies to determine an
appropriate "accident probability category" during the nsk analysis step described in Chapter
13 of this guide.
An important point to be made is that the analysis methods presented herein provide
their users with the annual accident probability expected for the entire area of concern and
not for specific locations or subregions within the locality Should it be desired to determine
the probability of a release near a specific populated area, a specific body of water, or a
specific environmentally sensitive habitat, it will be necessary for users to determine that
fraction of the overall community or facility exposure associated with this special area and
to adjust overall accident frequency predictions in an appropriate fashion
Special Notes
Wherever the term "spill" is used in the discussions that follow, readers are advised to
interpret the term as referring to any release or discharge of a hazardous material in a manner
capable of posing a threat to the public or the environment
The procedures that follow permit the user to estimate the number of "spills/year"
expected on average from individual activities or operations involving hazardous materials.
The average reoccurrence interval for a specific spill scenario can be determined by dividing
the number "1.0" by the predicted frequency of spills per year Thus, a spill frequency of 2 x
1O2 spills per year can be translated to mean that a spill can be expected to occur once every
11-2
-------
50 years on average. Where desired, readers may also sum the spill frequencies denved for
similar individual activities and then determine the reoccurrence interval associated with the
combination using the same methodology
Throughout this chapter, historical accident rates and other data are frequently
presented in "scientific notation" Readers unfamilar with this method of representing very
small (or very large) numbers are referred to Appendix A for guidance
11J GENERAL SEVERITY OF ACCIDENTS CONSIDERED
The types of accident scenarios that could theoretically be covered in this chapter range
from minor spills of gasoline at service stations to major catastrophes that occur once every
10 or 20 years on average on a worldwide basis. The primary focus, however, is somewhere
between these extremes, and on the types of events which occur somewhere in the United
States every week or month, but for which the nsks to any one specific community might be
low but nevertheless significant In other words, we are not highly concerned with routine
and common spills and discharges of relatively minor significance Nor are we overly
concerned with extremely rare events.
For a general perspective on hazardous materials accidents, consider that there are
many thousands of hazardous materials releases which occur each year in transportation and
at fixed facilities, yet the vast majority are of very limited consequence Tables 11 1 to 11 3
give some idea of the total numbers of accidents reported each year and the relative
contribution made by various activities. More specifically, Table 111 addresses the number
of evacuations in recent years that were of sufficient magnitude to warrant reporting. Table
11.2 provides an estimate of the total number of accidents over a ten year period involving
hazardous material releases from transportation activities, while Table 11.3 focuses on major
events by type of activity Although the data presented in the latter table are somewhat
dated, they are nevertheless sufficient to demonstrate that only a small fraction of all
accidents result in major consequences.
11.3 BULK TRANSPORTATION OF HAZARDOUS MATERIALS BY HIGHWAY
Bulk transportation of hazardous materials by highway involves the use of tank trucks
or trailers and certain types of more specialized bulk cargo vehicles In all, trucks transport
more than sixty percent of the hazardous materials not earned by pipelines, with just under
fifty percent of this material being gasoline (OTA, July 1986) Average trip lengths are 28
miles for gasoline trucks and 260 miles for chemical trucks, implying that most gasoline
shipments are very localized while chemical shipments may be regional or interstate Since
trucks carry hazardous materials the greatest number of miles near populated areas, and are
also responsible for the largest number of shipments, it is not surprising that this mode of
transportation is also responsible for the greatest number of accidents.
11-3
-------
TABLE 11.1
CHEMICAL ACCIDENTS REQUIRING EVACUATIONS
Type of Accident
Train derailment
Railcar spill/fire
Truck accident
Truck spill/fire
Chemical plant release
Industrial plant release
Pipeline
Ship incident
Waste site accident
Other*
TOTAL
1980-1984
55
23
35
32
43
78
4
4
7
14
295
1984
8
5
5
7
5
24
0
1
1
1
57
*Includes helicopter crash, plane crash, sewer gas episode, oil well explosion, swimming pool
chlorine accident, pesticide spill hi retail store, mine fire, two military missile silo accidents,
and two electrical transformer leaks.
Source:
Sorensen, 1986
TABLE 11.2
HAZARDOUS MATERIAL TRANSPORTATION ACCIDENTS
Mode
Highway
Rail
Air
Water
Number of Accidents
Per Year1
10,000-15,000
1,000
200
20
1973-1983 Average1
10,289
975
150
26
Source: Materials Transportation Bureau, 1983 and Check et al, 1985
*OTA, March, 1986
11-4
-------
TABLE 11.3
MAJOR HAZARDOUS MATERIALS ACCIDENTS*
Activity
Chemical plants
Oil refineries and storage
Gas storage tanks
Oil drilling ngs
Pipelines
Fireworks plants
Marine tankers, barges
Railroad tank cars
Trucks
TOTAL
1964-1973
6
10
1
2
1
0
8
5
3
36
1953-1973
12
13
2
2
1
2
15
8
8
63
*10 or more fatalities, 30 or more injuries, $3,000,000 or more
in property damage
Source Office of Radiation Programs, 1980
11-5
-------
Tank trucks are usually tractor-semitrailer vehicles or smaller bobtail-type units. The
tanks themselves are usually constructed of steel or an aluminum alloy, but may also be
stainless steel, nickel and other materials. Capacities are usually in the range of 3000-10,000
gallons, although slightly smaller and larger units are available. Intermodal tanks which
consist of a tank within a protective rigid framework, one-ton tanks which are lifted on and
off the transporting vehicle, and large gas cylinder bundles are also commonly used for bulk
transport by highway.
Commodity breakdowns for trucks, as described in various data sources are not
considered very accurate and vary widely. However, a one-month survey of cargoes in
Virginia found a fairly close match (by percentage) to the average distribution of commodi-
ties involved in accidents. The comparison for commodities involved in accidents in
Virginia also matched national accident breakdowns fairly well (Urbanek and Barber, 1980).
National involvement in accidents by type of commodity for the time period July
1973-December 1978 was:
Flammable liquids 60.5%
Combustible liquids 16.3%
Corrosives 11.6%
Flammable compressed gases 3.2%
Oxidizers 2.1%
Poisons (liquid or solid) 2.1%
Nonflammable compressed gases 1.9%
Explosives 1.5%
Radioactive materials 0.5%
Flammable solids 0.3%
Causes and Examples of Past Accidents
Truck accidents on roadways, regardless of the cargo involved, are generally due to:
• Collisions with other vehicles
• Collisions with fixed objects such as bridges or overpass supports
• Running off the road
• Overturns due to excessive speed on curves
These four events are most likely to result in a release of large quantities of hazardous
materials, and are predominantly the result of human error. Smaller releases may arise due to
defective or loose valves, fittings or couplings; weld failures; and various other structural
defects. (Note: Loading or unloading spills are considered separately in the category of fixed
facilities below and actually result in the majority of overall releases involving trucks)
11-6
-------
The severity and consequences of discharges resulting from truck accidents can vary
widely. Some examples include:
• A tank truck carrying 11,000 gallons of gasoline blew a tire, struck a cement
barrier which ripped open the side of the tank, and then burst into flames on
Interstate 95 near Peabody, Massachusetts. State troopers closed down the
highway while an emergency crew from Logan International Airport spread
foam on the wreckage. The highway remained partially closed for several
days, since one section had melted and needed to be replaced. There were no
injuries or deaths. (December 3,1985)
• Littleton, New Hampshire, was spared a potentially catastrophic accident
when a tank trailer loaded with 9,200 gallons of liquefied propane jack-knifed
on an icy hill and tipped on its side about 75 yards from a large storage tank
of liquid propane and less than 100 yards from large fuel oil storage tanks.
No propane leaked from the truck, but a diesel fuel tank was ruptured, 1500
people residing within a half mile radius were evacuated until the propane
was safely transferred to another vehicle the next day. (February 11,1982)
• A truck pulling two tank trailers loaded with molten sulphur collided with a
highway barrier on a toll bridge and burst into flames taking two lives and
injuring twenty-six Firemen encountered difficulty extinguishing the fire and
rescuing victims. Visibility at the time was poor due to fog and the spilled
sulfur had burned through water supply lines. (January 19,1986)
• In Houston, a tank truck carrying liquefied anhydrous ammonia collided with
a car and fell from an elevated highway to a busy freeway. The truck
exploded violently on impact releasing billowing clouds of ammonia Four
persons (including the truck driver) were killed, dozens of motorists were
overcome by the fumes in a three-mile area, and at least 100 were treated at
area hospitals. The vapors and fumes were so thick that police helicopters
were initially repelled. The city was forced to use all available ambulances
and pnvately-owned hearses to transport the injured. (May 12,1976)
• Although certain details are unclear, a tank truck carrying liquid propylene
sprang a leak in the vicinity of a crowded campsite in Spain Flammable
gases spread from the truck, encountered a source of ignition, flashed back to
the vehicle, and caused a BLEVE with a large fireball The death toll from
burns was approximately 170 Numerous other people suffered moderate to
severe burns but otherwise survived (July 11,1978)
11-7
-------
Suggested Approach for Assessment of Accident Potential
Since we are concerned with accidents with the potential to cause major problems for a
community or other jurisdiction and not those which are handled on a routine basis, it is best
to focus on vehicular accidents rather than relatively minor leaks from valves, fittings, or
open relief valves. Based on the information discussed in Appendix F, an average accident
rate of 2 x 106 accidents/mile is considered representative of the general expenence of trucks
carrying bulk quantities of hazardous materials If adequate local/state data are available for
determination of individual accident rates for divided and undivided roadways, then* use is
recommended because the resulting rates will more accurately reflect accident probabilities
under local conditions.
With respect to the fraction of truck accidents that result in a spill or discharge, the
available data suggest a consensus opinion on the order of 0.50 (50%) if all spills including
very minor valve and fitting leaks are considered Omitting these, a spill appears to result
from an accident in about 0.15-0 20 (15% - 20%) of accidents. A value of 0.20 (20%) is
therefore adopted for use for the sake of conservatism.
Based upon available spill size distributions, and considering the likely causes of
accidents, the following distribution is suggested for general use:
• 10% cargo loss (thru 1" hole) or 1000 gal --60% of the time
30% cargo loss (thru 2" hole) or 3000 gal — 20% of the time
• 100% cargo loss (instantly) or 10,000 gal --20% of the time
These values cover the range of significant releases If desired, a two-point distribution
assuming that 3000-gallon spills occur 80% of the time and 10,000-gallon spills occur 20%
of the time may be used to simplify consequence analysis procedures.
The suggested accident rates and other factors for truck accidents are summarized in
Table 11.4. Worksheet 11.1 presents a simplified format for computing the annual average
probabilities of truck accidents resulting in spills of various amounts A copy of the
worksheet should be completed for each hazardous material transported in bulk by truck
within or through the community, or (if desired) groups of chemicals or materials posing
similar threats in the event of a release may be analyzed together Local information needed
for the task includes:
• Matenal(s) of concern
• Annual number of shipments
• Total capacity per shipment
11-8
-------
TABLE 11.4
SUGGESTED FIGURES FOR TRUCK TRANSPORTATION
Accident Rate: 2 x 106 accidents/mile
Conditional Spill Probability 0 2 for significant spills
Spill Size Distribution 0 60 for 10% loss of cargo through 1" hole, or 1,000
gal
0 20 for 30% of cargo through 2" hole, or 3,000 gal
0.20 for 100% of cargo, or 10,000 gal
Note: Worksheet 11.1 demonstrates how these figures can be used to estimate annual
accident probabilities and associated spill volumes for truck transportation.
11-9
-------
WORKSHEET 11.1
ESTIMATING BULK TRUCK TRANSPORTATION RELEASE FREQUENCIES
Hazardous Material(s):
Total Number of Annual Shipments:
Length of Route of Concern:
Total Number of Miles Per Year*:
A =
(loaded trucks only)
Tj _ m
(miles within jurisdiction)
= AxB=-
Accident Frequency:
Spill Frequency:
= Cx2xlO« =
= Dx0.2 =
_(accidents/year)
_(spills/year)
Spills by Size*
10% loss of cargo (1" hole) or 1,000 gal:
30% loss of cargo (2" hole) or 3,000 gal
Ex06 =
E x 0.2 =
100% loss of cargo or 10,000 gal-
Ex02 =
Notes:
_(spills/year)
_(spills/year)
_(spills/year)
*If there are a number of different routes with varying numbers of shipments, multiply the
number of shipments by the route length for each route individually and then sum For
example, 100 shipments of 15 miles and 50 shipments of 7 miles would give (100 x 15) +
(50 x 7) = 1850 total miles.
+ The user may consider all three scenarios for consequence modelling and planning
purposes or just the largest spill.
11-10
-------
• Total length of route within community
• Type of roadways travelled (if specific rates by type of highway are available)
Note that it may be necessary in some cases to combine the number of shipments with the
length of route within the community to compute total mileage because there may be several
routes for the same material. A prune example is for gasoline which may be delivered to
several locations within any community or jurisdiction as well as driven through the locality.
Worksheet 11.1 provides further information on how this may be accomplished.
Additional Data/Methodologies
Should more precise methodologies be desired or should special circumstances warrant
attention, the reader can consider one of the techniques listed below. First efforts should be
directed at obtaining more precise data either on a local, county, state or regional basis. The
data specific to one carrier should not be broadly applied, however, as there can be major
differences between earners - even when they operate in the same area. This may occur due
to differences in truck design and upkeep, characteristics of the cargo, training of the drivers,
enforcement of speed restrictions, and other factors.
More detailed methodologies which take into account specific accident situations
include:
• Separate models to predict accidents on interstates, rural highways or urban
arterials as a function of several input variables (Urbanek and Barber, 1980).
• Analyses of rail/highway grade crossing accidents (National Transportation
Safety Board, 1981). Note: There are only 60 or so of these each year,
nationwide, on average. They usually involve trucks carrying petroleum
products and occur close to distribution/storage terminals, with very localized
impacts.
• Breakdowns of rates by rural, urban, suburban and number of lanes (Smith
and Wilmot, 1982).
• Risk assessments of transportation through tunnels (Considine, 1986).
• Detailed considerations of the severities of various types of accidents for
particular vehicle configurations (Clarke et al, 1976).
11-11
-------
Use of Results in Consequence Analysis
Each of the three accident scenarios denoted and evaluated above actually provides the
user with two options, the first being assumption of a certain percentage of cargo loss
through an orifice of a given size, and the second being assumption of a specific total amount
of cargo loss. For example, in the first scenario considered, the user can assume that 10% of
the total cargo of the vehicle is discharged through a hole having a diameter of one inch, or
alternatively, simply assume that 1000 gallons of presumably liquid cargo are released before
the discharge is terminated for one reason or another under average accident conditions
These results are based on generalized historical records of past accidents and provide one
way in which the ultimate consequences of an incident can be evaluated by use of the
analysis procedures discussed in Chapter 12 of this guide Conversely, depending upon the
type of hazardous material involved and the desires of the user, these scenarios can be further
refined for consequence analysis purposes, taking better advantage of locally available
information.
Where the user wishes to assume a fixed percentage of cargo loss (this requiring
knowledge of total cargo amounts) or a fixed amount of spillage for liquid cargos, the spill
can be assumed to take place instantaneously, thus obviating need for use of discharge rate
and duration estimation methods discussed in Chapter 12 and available in the computer
program provided with this guide It is cautioned, however, that the assumption of an
instantaneous release in such situations may greatly overestimate resulting hazard zones from
evolution of toxic or flammable vapors, fires, or explosions Spill amounts presented in units
of gallons of liquid can be converted to the units of pounds required by the computer
program by use of the following expression:
Amount in pounds = 8.34 x Amount in gallons x Liquid specific gravity
Where a hole size is specified and cargo tank or compartment dimensions are known, it
is alternatively possible (and recommended) to utilize available discharge rate and duration
estimation procedures to obtain an ultimately more realistic indication of accident conse-
quences. This is particularly advisable when the cargo is a compressed gas, a liquefied
compressed gas, or an otherwise highly volatile material. The suggested percentage of cargo
loss, when less than 100%, can be ignored, if necessary, in deference to the results obtained
from the discharge models. (This is due to the fact that the hazard zone will be primarily
determined by the release rate and the first ten or so minutes of the release; the ultimate
release quantity is less significant)
11-12
-------
11.4 BULK TRANSPORTATION HAZARDOUS MATERIALS BY RAIL
According to recent statistics, about 80 million tons of hazardous materials are shipped
annually by rail in the U.S. (OTA, March 1986) The majority of these shipments are in a
single tank permanently mounted on a rail car. Exceptions include multi-tank tank cars (the
units are usually ton containers), seamless steel cylinders (as for very high pressure service),
and compartmented tank cars in which each compartment is treated as a separate tank. The
sizes of these will range from a few hundred gallons in the case of a ton container to 45,000
gallons in so-called jumbo tank cars. Since 1970, however, the capacity of new tank cars has
been limited to 34,500 gallons There is also occasional use of mtermodal tanks, as
mentioned under truck transportation.
The design, construction, inspection, and use of tank cars are regulated by both the
American Association of Railroads (AAR) and the Federal Railroad Administration (FRA)
within the U.S. Department of Transportation (USDOT) Carbon steel is used to construct
over 90% of the tanks, with aluminum for most but not all of the remainder Nickel or nickel
alloy is found in acid service, and there are a small number of stainless steel cars (Wright and
Student, 1985). Safety relief valves (and vents) are required, unless otherwise specified
Cars are usually classified into the categones of pressure tank cars, non-pressure tank cars,
cryogenic liquid tank cars, and miscellaneous tank cars. Tanks may be lined, insulated, and
possibly fitted with heating coils Some may have special thermal protection to prevent
BLEVEs or other explosions in the event of exposures to pool fires or flame jets Relatively
recent regulations have required shelf couplers - which limit potential for the puncturing
adjacent cars in the event of a derailment or collision - for all new and old cars Cars
carrying liquefied flammable gases or ammonia have been required to have head shields to
further limit puncture potential, and new housings for bottom outlets have also been adopted
It has been estimated that 35% of all freight trains carry hazardous materials, but that
only 7 5% of railroad accidents involve trains carrying these materials (von Herberg, 1979)
This figure corresponds to the percentage of all cars which carry chemicals and allied
products or petroleum products versus the total number of cars on an annual basis (AAR,
1985).
A fairly extensive data base on commodity flows is available for railroads The July
1986 OTA report provides the following breakdown on a tonnage basis*
Flammable liquids 26%
Corrosive matenals 25%
Combustible liquids/
nonflammable compressed gases 22% (less than 12% each)
Flammable compressed gases 12%
11-13
-------
Poison B 3%
Poison A 0.1%
Radioactive materials 0.03%
Detailed breakdowns by state, railroad company, and other divisions are also available from
various sources described in the OTA report.
Data on rail yards shows that the number of hazardous materials cars handled by each
ranges from 1-15 percent of the total throughput (Chemical and Engineering News, July 29,
1985). In one study, based on data through 1977, it was found that 36% of derailments and
73% of collisions occurred in such yards (Nayak et al, 1983).
Causes and Examples of Past Accidents
For releases of hazardous materials from rail accidents, there are two types of events of
concern. The most important for this analysis is the accident that involves a collision or a
derailment, since these typically involve the largest spills or discharges. However, there is a
second class of releases which may arise from fitting or seal leaks, relief valve leaks, and
other releases associated with improper tightening of closures or defective equipment
Harvey et al (1987) estimate that these account for 70% of the roughly 1000 releases each
year. Rail accidents, like those for trucks, can result in virtually no adverse consequences up
to very large losses of life, depending on the materials involved and the circumstances of the
accident Many of the more severe accidents occur in yards and on sidings (Wolfe, 1984)
As for truck transportation, incidents arising during loading or unloading operations are
addressed under fixed facilities.
AAR data (Wolfe, 1984) showed that the materials most often involved in accidents
with more than $100,000 of damage in 1981 were:
LPG
• Acrylonitrile
• Fuel oil
• Vinyl chloride
• Anhydrous ammonia
The same source found that there were less than 30 accidents each year with this level of
damage.
Examples of past accidents involving the rail transportation of hazardous materials are
given below. These particular incidents include some of the most severe that have occurred
in recent times.
11-14
-------
In Waverly, Tennessee the derailment of two propane cars was treated rather
casually as crews worked to clear the track and nght the cars. As this was
being done, some 40 hours after the derailment, propane began leaking from
one car, reached a source of ignition, flashed back to the car, and caused an
explosion and fireball. The town center looked like a battle scene after the
explosion with 16 dead, 54 requiring hospitalization for burns, and 42
requiring outpatient care (February 24,1978)
Over 240,000 people were evacuated for all or part of a week in Mississauga,
Canada after a derailment involving 11 propane cars, 1 chlonne car and 10
cars loaded with other chemicals. The wreckage produced a series of
explosions, launched missiles more than half a mile, and prompted fears of a
massive chlonne release. No fatalities or major injuries occurred partly due
to the quick accident response of authorities, a well executed emergency
evacuation plan, and various fortuitous circumstances. (November 10,1979)
A white cloud of toxic smoke towered a thousand feet over the community of
Miamisburg, Ohio and covered an area about a mile wide and 10-15 miles
long at one point after a derailment caused a car containing white phosphorus
to fail and ignite. The intense heat and difficulties in controlling the fire
forced authorities to wait four days for the blaze to burn itself out. Eleven
persons were hospitalized after exposure to the toxic smoke, with a total of
273 being treated for skin, eye, and lung irritations. The 40,000 (or more)
evacuees were the largest number ever resulting from a train accident in the
U.S. (July 8,1986)
Suggested Approach for Assessment of Accident Potential
Based on the data presented in Appendix F, it is suggested that an accident rate of 3 x
per train-mile be used for mainline track. To convert this to a per car-mile basis, it is
assumed that 0.20 (20%) of the cars will be damaged in an accident (based upon data
presented in Nayak, 1979). The overall rate therefore becomes 0 2 x 3 x 10V train-miles or
about 6 x 10-7 per car-mile.
The accident rate for rail yards is obtained by taking 1 3 x 10-s accidents per train-mile
and a 20% damage estimate to obtain about 3 x lO-Ycar-mile for the track m yards. Sidings
also pose a risk, but these nsks are considered herein to be overshadowed by those associated
with mainline and yard track.
It is suggested that 0.15 (15%) of accidents be assumed as resulting in a spill for both
mainline and yard accidents, as no data are available to permit distinctions between these
events.
11-15
-------
With respect to the distributions of spill amounts m accidents, the available data
suggest use of:
• 3,000 gallons or 10% of cargo (thru 2" hole) -50% of the time
10,000 gallons or 30% of cargo (thru 2" hole) - 20% of the time
• 30,000 gallons or 100% loss of cargo - 30% of the time
The higher weighting of the last category partially accounts for the potential for more than
one car to release part of its contents in an accident.
Table 11.5 summarizes the accident rates and other factors suggested for use, while
Worksheet 11.2 outlines the procedure for determining the annual average probability of an
accident involving spills of various amounts. A copy of the worksheet should be completed
for each hazardous material transported in bulk by rail within or through the community (As
for trucks, groups of chemicals posing similar hazards may be considered together) Local
information that will be required to accomplish the effort includes'
• Matenal(s) of concern
• Annual number of cars
• Total capacity per car
• Total miles of mainline track within community
• Total miles of yard track travelled by a typical car
Railcar trips and associated mileage involving loaded vehicles are of primary concern as
these pose the greatest risk Nevertheless, it is important to realize that the residual materials
within tank vehicles considered "empty" can also pose a hazard under certain circumstances
If local data on accident rates or spill frequencies are available, they can be directly
substituted for the rates and other factors listed in Table 115.
Additional Data/Methodologies
Should a more detailed evaluation be desired or required, readers can consider use one
of the techniques listed below. However, any effort to improve the specificity of accident
predictions should probably first involve the determination and use of individual state or
railroad company accident rates for specific routes.
More detailed evaluations of rail transportation can also include consideration of
several different factors (alone or in combination) in the analysis. Examples of these include:
• Detailed review of accident seventy to determine the overall likelihood of
puncture, crush, impact and fire scenarios (Clarke et al, 1976)
11-16
-------
TABLE 11.5
SUGGESTED FIGURES FOR RAIL TRANSPORTATION
Mainline accident: 6 x lO7/car-mile
Yard accident rate: 3 x KWcar-mile
Spill size distribution: 0.5 for 10% cargo loss through a 2" hole, or
3,000 gallons
0 2 for 30% cargo loss through a 2" hole, or
10,000 gallons
0 3 for complete loss of a cargo load, or about
30,000 gallons on average
Note: Worksheet 11.2 demonstrates how these figures can be used to estimate annual
accident probabilities and associated spill probabilities for rail transportation
11-17
-------
WORKSHEET 11.2
ESTIMATING BULK RAIL TRANSPORTATION RELEASE FREQUENCIES
Hazardous Material(s):
Number of Cars Per Year:
Number of Car-Miles in Yards:
Number of Car-Miles on Mainline:
A=-
(loaded cars only)
B=-
(miles per trip in jurisdiction)
C=-
(miles per top in jurisdiction)
Accident Frequency: D = (AxBx3x 10«) + (A x C x 6 x 1O7) =
_(accidents/year)
Spill Frequency:
= Dx015 =
_(spills/year)
Spills by Size*
10% loss of cargo (2" hole) or 3,000 gal: E x 0 5 =
30% loss of cargo (2" hole) or 10,000 gal: E x 0.2 =
_(spills/year)
_(spills/year)
100% loss of cargo or 30,000 gal:
Ex03 =
_(spills/year)
Notes:
*The user may consider all three scenarios for consequence modelling and planning
purposes or j'ust the largest spill
11-18
-------
" WA*
• Detailed consideration of specific types of accidents and associated leak sizes.
(Note One study in Finland considered 14 types of accidents and four
categones of leaks, including valve leaks, broken valves, moderate breaks and
punctures, and large breaks) (Lautkaski et al, 1979)
• Separate consideration of different classes or quality of track (FRA, 1988,
gives some the additional information needed for this type of evaluation)
• Separate consideration of mainline, yard, and sidings
Use of Results in Consequence Analysis
Each of the three accident scenarios denoted and evaluated above actually provides the
user with two options, the first being assumption of a certain percentage of cargo loss
through an orifice of a given size, and the second being assumption of a specific total amount
of cargo loss For example, in the first scenario considered, the user can assume that 10% of
the total cargo of the vehicle is discharged through a hole having a diameter of two inches, or
alternatively, simply assume that 3000 gallons of presumably liquid cargo are released before
the discharge is terminated for one reason or another under average accident conditions
These results are based on generalized historical records of past accidents and provide one
way in which the ultimate consequences of an incident can be evaluated by use of the
analysis procedures discussed in Chapter 12 of this guide. Conversely, depending upon the
type of hazardous material involved and the desires of the user, these scenarios can be further
refined for consequence analysis purposes, taking better advantage of locally available
information.
Where the user wishes to assume a fixed percentage of cargo loss (this requiring
knowledge of total cargo amounts) or a fixed amount of spillage for liquid cargos, the spill
can be assumed to take place instantaneously, thus obviating need for use of discharge rate
and duration estimation methods discussed in Chapter 12 and available in the computer
program provided with this guide. It is cautioned, however, that the assumption of an
instantaneous release in such situations may greatly overestimate resulting hazard zones from
evolution of toxic or flammable vapors, fires, or explosions Spill amounts presented in units
of gallons of liquid can be converted to the units of pounds required by the computer
program by use of the following expression.
Amount in pounds = 8.34 x Amount in gallons x Liquid specific gravity
Where a hole size is specified and cargo tank or compartment dimensions are known, it
is alternatively possible (and recommended) to utilize available discharge rate and duration
estimation procedures to obtain an ultimately more realistic indication of accident conse-
quences This is particularly advisable when the cargo is a compressed gas, a liquefied
11-19
-------
compressed gas, or an otherwise highly volatile material. The suggested percentage of cargo
loss, when less than 100%, can be ignored, if necessary, in deference to the results obtained
from the discharge models.
113 BULK TRANSPORTATION OF HAZARDOUS MATERIALS BY MARINE VES-
SELS
A large portion (about 550 million tons in 1982 according to the OTA) of the
hazardous materials shipped annually in the United States is transported by barge or other
marine vessel on coastal and inland waterways. Extensive regulations cover safety proce-
dures, cargo documentation, vessel construction and certification, hazardous material
transfers, and the handling of explosives or dangerous cargos within or near waterfront
facilities. Individual shipments can be vastly larger than those conveyed by rail or truck due
to the size differences among these conveyences.
The primary types of marine vessels used for bulk transportation of hazardous goods
are bulk liquefied gas earners, chemical tankers, oil tankers and tank barges, but bulk cargos
may also be found in smaller tanks placed on the decks of vessels or in standard intermodal
cargo containers. Some barges are self-propelled but most are designed to be pushed by a
tugboat singly or in arrays called "tows". Marine transportation generally involves volumes
of 300-600 thousand gallons in barges, though some such vessels are of far larger capacity.
Tank ships can have capacities that are ten times or more greater (OTA, July 1986).
Commodity flow data are compiled fairly rigorously for marine transportation Crude
petroleum, petroleum products, and chemicals and liquefied gases constitute a large fraction
of all shipments in and out of most major ports. Petroleum products include alcohols, crude
oil, refined fuels, solvents and residuals. Typical chemicals include sulfunc acid, benzene,
toluene, sodium hydroxide, inorganic speciality products, and fertilizers. The most frequent-
ly transported liquefied gases are propane and butane, but anhydrous ammonia, chlorine,
propylene, butylene and butadiene are also frequently transported by water, as are many
other bulk commodities
Causes and Examples of Past Accidents
Marine transportation is generally at slow speeds and involves many precautions and
traffic controls Hence, it has the lowest accident rate per ton-mile and the lowest number of
accidents. However, the large energies involved when these massive vessels strike each
other or other objects can result in severe consequences at times in terms of cargo loss As
with other modes of transport, small releases can result from problems with gaskets, flanges,
valves or even the tanks themselves. The separation from population limits the consequences
of small releases, however.
11-20
-------
A few examples of accidents involving marine transportation of hazardous materials
include:
• A total of 23,000 gallons of leaded and unleaded gasoline spilled into the
southern approach to Cape Cod Canal when a barge ran aground. The Army
Corps of Engineers and the Coast Guard responded with divers and marine
safety personnel to assess the damage after the tugboat called for assistance.
The Coast Guard expected the sea current to dissipate the spill Environmen-
tal damage was said to be minimal. There were no deaths or injuries
(August 18,1986)
• A collision in the Houston Ship Channel between a tug and barges and a grain
ship resulted in an explosion and fire involving one 33,000 gallon tank of
butadiene. Two burning barges were towed to sea where they burned for five
days. (August?, 1980)
• A 565-foot long tanker carrying two million gallons of gasoline rammed an
unlighted oil drilling ng in the Gulf of Mexico. The tanker caught fire and
had to be abandoned. (August 21,1980).
• A barge carrying 400,000 gallons of acrylonitnle struck the Galveston
Causeway railroad bridge and ignited Resultant explosion caused one end of
the barge to sink and release an unknown amount of chemical into the water
(January 3,1982)
• Up to 40,000 pounds of hydrobromic acid spilled into the Mississippi
waterway after a collision between two ships. Violent reactions between the
acid and water required evacuation of 3000 people on the adjacent shore
(July 22,1980)
Suggested Approach for Assessment of Accident Potential
Based upon the information and data presented in Appendix F, and given the
understanding that harbors and inland waterbodies are of greater concern than shipping
activities on the open ocean and/or otherwise distant from coastlines (in terms of the
distances typically associated with spill effects that may pose a threat to human life and
health), accident rates and other spill characterization factors are presented below for.
• Collisions in lakes, nvers, and intercoastal waterways
• Groundings in lakes, nvers, and intercoastal waterways
• Collisions and groundings in harbors and bays
• Collisions/casualties while vessels are moored/docked
11-21
-------
An accident rate of lOVmile of travel is suggested for use for collisions in the first
category to cover both the lower expected accident rates on certain slow speed waterways
and the higher ones for congested, highly utilized routes. Based on Gulf Intercoastal
Waterway (ICWW) statistics, a grounding casualty rate of 5 x 1O6/ mile is suggested for the
serious type of grounding which could lead to a release. Note that this is also a "per mile"
rate. The harbor/bay area grounding and collision rate given below is "per transit", while the
moored collision rate is "per port call." (There are two transits per port call) The suggested
rate for groundings and collisions m a harbor area is lOVtransit, while the suggested casualty
rate for moored or docked vessels is 2 x 104 per port call.
If no distinction is being made with regard to vessel type and construction, it should be
assumed that 0.15 (15%) of accidents result in actual loss of cargo to the environment
Alternatively, it can be assumed that accidents involving single-hulled vessels result in cargo
loss 0.25 (25%) of the time and that accidents involving double-hulled and bottomed
watercraft result in cargo loss 0.05 (5%) of the time.
The recommended distribution of spill amounts is:
• 10% loss of cargo in one tank/compartment ~ 35% of the time
• 30% loss of cargo in one tank/compartment — 35% of the time
• Full loss of cargo in one tank/compartment — 30% of the time
This distribution is weighted toward more severe events than the spill distributions presented
earlier, because the earlier distributions are heavily influenced by minor fitting leaks.
Table 11.6 summarizes the accident rates and other factors suggested for use, while
Worksheet 11.3 outlines the procedure for determining the annual average probability of an
accident involving spills of various amounts. A copy of the worksheet should be completed
for each hazardous material or group of similar materials transported hi bulk by waterborne
vessels through the community or other jurisdiction of concern. Local information that will
be required to accomplish the effort includes:
• Material(s) of concern
• Maximum tank capacity of vessels carrying this material
• Total number of lake, river, or mtercoastal waterway miles in the area,
• Total ships traveling this route in a year,
• Total, cargo barges/tankers entenng and exiting the bay area or harbor,
• Total barge/tanker port calls,
11-22
-------
TABLE 11.6
SUGGESTED FIGURES FOR MARINE TRANSPORTATION
Accident Rates* 10-5/rmle for collision on lakes, rivers and inter-
coastal waterways
5 x IfrYmile for groundings on same
lOVtransit for collisions and groundings in
harbors/bays
2 x lOVport call for collisions/casualties while
moored
Conditional Spill Probabilities' 0 15 if using one rate regardless of vessel
0 05 for double-hulled/double-bottomed vessels
0.25 for single-hulled vessels
Spill Size Distribution: 0.35 for 10% loss of one tank or compartment
0.35 for 30% loss of one tank or compartment
0.30 for 100% loss of one tank or compartment
Note: Worksheet 11.3 demonstrates how these figures can be used to estimate annual
accident probabilities and associated spill probabilities for marine transportation.
11-23
-------
WORKSHEET 11.3
ESTIMATING BULK MARINE TRANSPORTATION RELEASE FREQUENCIES
Hazardous Material(s):
Length of Lake, River, ICWW
Route*:
Annual Number of Trips on Route*:
Annual Number of Transits of Har-
bor/Bay*:
Annual Number of Dockings*:
B=-
C=-
D=-
(miles within jurisdiction)
(loaded trips only)
(loaded transits only)
(loaded dockings only)
Accident Frequency: E = (AxBxl.5x 1O5) + (C x 1O3) + (D x 2 x 1O»)=•
(accidents/year)
Spill Frequency*:
F = Ex0.15=.
(all vessels)
OR
F = Ex0.25=.
(single hull)
_(spills/year)
_(spills/year)
F = Ex0.05 =
(double hull) '
_(spills/year)
Spills by Size°
10% loss of one tank or compartment:
30% loss of one tank or compartment:
100% loss of one tank or compartment:
Fx035 =
Fx0.35 =
Fx03 =
(spills/year)
(spills/year)
(spills/year)
Notes:
*If applicable
*If it is known how many vessels are single-hulled and how many are double hulled, this
worksheet can be completed twice; the first tome for single-hulled vessels and the second for
double-hulled.
0 The user may consider all three scenarios for consequence modelling and planning purposes
or just the largest spill.
11-24
-------
• Total river miles in the area,
• Number of barges/tankers traveling the river route in a year.
When totaling barge/tanker port calls or harbor transits, remember to count only those
involving actual carnage of hazardous materials. Empty vessels may pose some risk of fire
or explosion, but are not as important as loaded vessels. Also, note that all of the data listed
above may not be needed for every location. For example, a community located on a nver
and not having a harbor or bay area only needs the total nver miles and number of ships
traveling the nver. In this case, only moving collisions and nver groundings would be of
interest.
Nonimpact casualties, such as fire/explosions, hull and machinery damage or break-
downs, and structural failture have a very low likelihood of occurrence It is not considered
necessary to include them in the analysis
Additional Data/Methodologies
A more detailed and accurate analysis may be performed if one chooses to denve the
appropnate casualty rates and conditional probabilities for a specific harbor or water route
using local data. An accident rate may be derived by combining a measure of vessel
movements with the number of past accidents reported for the type of movement being
looked at The movement measure may consist of' 1) the total waterbody, nver, or waterway
miles traveled; 2) the total port calls made; or 3) the total harbor transits Counts should be
made only for loaded hazardous material tankers and barges The number of accidents is
then divided by the total number of transits or miles, as appropnate, for a similar penod of
time to determine the accident rate.
Conditional spill probabilities and spill distnbutions are denved in a similar fashion
The probability of a spill is obtained by dividing the number of spill accidents by the total
number of accidents that occurred for a specific type of movement A spill amount
distnbution from local or national data can then be applied to determine the percentage of the
spills expected in vanous size ranges.
Several other means exist for estimating the probability of marine casualties and spills
but they are generally highly technical, time-consuming, and occasionally quite expensive.
The reader is directed to the reference list at the end of the chapter for further information
11-25
-------
Use of Results in Consequence Analysis
Due to the special nature of the discharges from marine vessels, and difficulties in
estimating discharge rates and durations when there is water outside the cargo tank rather
than air, it is recommended (in the absence of more specific information or assumptions) that
all discharges be assumed to take place instantaneously for emergency planning purposes.
Spill amounts available in units of gallons of liquid can be converted to units of pounds by
use of the following expression-
Amount in pounds = 8.34 x Amount in gallons x Liquid specific gravity
11.6 TRANSPORTATION OF HAZARDOUS MATERIALS BY PIPELINE
Pipelines in the United States primarily carry petroleum liquids, such as crude oil,
gasoline and natural gas liquids, and energy gases which include natural gas and liquefied
petroleum gas (LPG). To a much smaller extent, pipelines also transport ethane, ethylene,
liquefied natural gas (LNG), anhydrous ammonia, carbon monoxide, sour (hydrogen sulfide
containing) gas, and many other chemicals. The majority of these pipelines are between a
limited number of suppliers and users, as opposed to natural gas transmission lines. Low
pressure gas distribution lines found within many cities and towns are not the focus of this
section.
Pipelines are generally constructed out of steel, although some cast iron is still in use
and plastic, nickel alloys, stainless steel, carbon steel and other materials are also used.
Diameters also vary tremendously, from 2 to 4 inches to 36 inches and over. The more
hazardous materials tend to be conveyed in lines at the smaller end of the size range
Pressures also span a wide range, and can be several thousands of pounds per square inch or
more in the highest pressure lines encountered.
In order to reduce failures caused by corrosion, pipelines are frequently coated and/or
cathodically protected. Lines may also be insulated, heated, double-piped for additional
protection and control, and protected with leak detection and shutdown systems.
Causes and Examples of Past Accidents
Pipeline failures may be a result of
• Internal corrosion -- especially on two-phase flow lines and those in sour
service
• External corrosion ~ from defects in protective systems, in cased crossings
beneath roads and railway lines
11-26
-------
• External impact ~ due to farm or construction machinery
• Structural failures or mechanical defects - as a result of defective seams or
welds
• Natural hazards — from seismic events, subsidence, etc.
Operating errors and construction defects are also potential causes of pipeline incidents.
Leaks may also occur at valves and pump stations.
While there are over a thousand leaks reported each year, many of these are very small
and have minor or no consequences to the public For instance, 1984 data show the
following number of incidents and deaths resulting from pipeline failures (Transportation
Systems Center, 1985):
Gas pipelines
Liquid pipelines
No. of failures
967
188
No. of deaths
35
0
These rates are not particularly consistent from year to year. In 1983, there were 1575 gas
pipeline failures, but only 12 deaths resulted There were also 6 deaths from liquid pipeline
failures.
The following three examples illustrate typical accidents involving pipeline transport of
propane and natural gas No injuries occurred in these accidents because of their remote
locations.
• In Port Hudson, Missouri, propane escaped from a pipeline, flowed into a
sparsely inhabited valley, ignited, and exploded with a blast equivalent to 50
tons of TNT. No fatalities or injuries occurred because the area was so
sparsely inhabited. (September 12,1977)
• In Prattville, Alabama, a natural gas pipeline exploded into two fireballs, and
shot flames 600 feet high. Two houses were scorched; 200 people were
evacuated, but no one was hurt. (My 12,1986)
• A natural gas pipeline explosion ignited a 200-ft flame jet and left a huge
crater: 60 ft across and 20 ft deep. No one was injured but an unmanned
metering station at the accident site was destroyed. (November 2,1985)
11-27
-------
Another typical pipeline accident did not have so fortunate an outcome as the above
three.
• In Jackson, Louisiana, a natural gas pipeline explosion resulted in 50-foot
high flames. The blaze was battled by firefighters for several hours. The five
dead and twenty injured were gas company employees repairing the pipeline.
(November 26,1984)
Suggested Approach for Assessment of Accident Potential
Based on the information presented in Appendix F, an accident rate of 1.5 x lOVmi-yr
is suggested for lines of unknown size or lines less than 20" in diameter For pipelines with
diameters greater than or equal to 20", a rate of 5 x 10Vmi-yr is proposed.
The following spill size distribution, incorporating the limited data available , is
suggested for analyzing pipeline releases of hazardous materials*
• For liquid pipelines: discharge computed using consequence analysis proce-
dures of Chapter 12 assuming a complete line break along the route of the
pipeline — 20% of the time
• For gas pipelines: discharge computed using consequence analysis proce-
dures of Chapter 12 assuming complete line break along the route of the
pipeline — 20% of the time
• For either gas or liquid pipelines. 1 hour release through 1" hole 80% of
the tune
Table 11.7 summarizes these rates, and Worksheet 11.4 demonstrates the procedures for their
use.
The application of this material requires local information on:
• Material of concern
• Length of pipeline within jurisdiction
• Pipeline diameter
• Flow rate (capacity) of pipeline
• Presence (or not) of a leak detection and emergency shutdown system
Should a pipeline be a very short segment between two facilities, it is possible to include it
with one of the facilities, rather than analyzing it separately.
11-28
-------
TABLE 11.7
SUGGESTED FIGURES FOR PIPELINE TRANSPORTATION
Accident Rates: 1.5 x 10 Vmi-yr with diameters less than 20"
of if diameter is not known
5 0 x lO-Vmi-yr lines with diameters greater
than or equal to 20"
Spill Size Distribution: 0.20 for 15 mm. (or 1 hour if no emergency
shutdown) at the capacity flow rate through
an orifice equal to the pipe size
0.80 for 1 hour release through 1" hole
Note: Worksheet 11.4 provides guidance on how to utilize these rates and probabilities
for estimating releases of hazardous materials from pipelines.
11-29
-------
WORKSHEET 11.4
ESTIMATING PIPELINE RELEASE FREQUENCIES
Hazardous Material(s):
Length of Pipeline of Unspecified A = •
Diameter: (miles in jurisdiction)
Length of Pipeline < 20" in Diameter: B =
(miles in jurisdiction)
Length of Pipeline ^ 20" in Diameter: C =
(miles in jurisdiction)
Spill Frequency: D = (A x 1.5 x 1O') + (B x 1.5 x 10*) + (C x 5 x 1(H) = (spills/year)
Spills by Size*
1 hour release through 1" hole: D x 0.8 = (spills/year)
Complete line break calculated according D x 0.2 = _(spills/year)
to procedures given in Chapter 12:
Note:
*The user may consider both scenarios for consequence modelling and planning purposes
or just the scenario posing the greatest threat to public safety. If several pipelines have
been grouped by diameter, the line posing the greatest threat should not automatically be
assumed to be the line with the largest diameter within any group. Rather, the various
pipelines should be individually evaluated using the consequence analysis procedures
described in Chapter 12 to determine the actual worst case scenario, if this is desired.
11-30
-------
Additional Data/Methodologies
The main alternative to the appioach given above is the specific consideration of
pipeline design and the operating environment to more precisely determine the failure rate
associated with each potential cause of failure (Techmca, 1983).
Use of Results in Consequence Analysis
For pipelines conveying compressed (but not liquefied gas), the consequence analysis
procedures described in Chapter 12 and incorporated into its associated computer program
are fully capable of estimating the rate and duration of gas release from either full line breaks
or smaller leaks.
Due to the complexity of the problem, the computer program's liquid pipeline
discharge model is only capable of addressing full line breaks It will be necessary to consult
with the pipeline owner or operator for assistance in estimating discharge rates and durations
for outflows from one inch diameter holes if these scenarios are considered worthy of
analysis.
11.7 HANDLING AND TRANSFER OF HAZARDOUS MATERIALS AT FIXED
FACILITIES
A broad range of facilities may pose potential risks associated with the release of
hazardous materials These can include, large refineries, chemical plants, and storage
terminals, more moderately sized industrial users, warehouses, and isolated storage tanks for
water treatment, small quantity users/storage as may be found in high school and college
laboratones, florists, greenhouses, hardware and automotive stores, paint stores, etc.
As a result of this broad spectrum of potential spill sources, virtually no accurate data
exists to determine the magnitude of this problem Marine terminals and loading/unloading
activities for rail cars and trucks are somewhat more widely reported and are considered
within this overall category.
FEMA has a database that identifies the number of chemical and petroleum facilities by
county. These facilities are broken down into chemical and allied products, petroleum and
coal products, and rubber and miscellaneous plastic products No accident information is
maintained, however As of 1981, there were 16,000 chemical process industry plants with
20 or more employees and 6,000 plants with 100 or more employees Los Angeles and Cook
(Illinois) counties each have over 200 plants The number of counties with between 11 and
50 plants is 160, all the rest have ten or fewer plants About 50% of all the counties m the
U.S. have no chemical process industry plants (Check et al, 1985) (This is not to say that
11-31
-------
there are no facilities handling hazardous materials in these counties.) McGraw Hill has
published a map of the plant locations and a booklet entitled "Census of the Chemical
Process Industries."
Causes and Examples of Past Accidents
Releases from fixed facilities may arise from storage tank or container ruptures or
leaks, piping ruptures or leaks, releases through safety and relief valves, fire-induced
releases, other equipment failures, malicious or deliberate actions, overfills and overflows of
storage tanks, human errors, open valves, failed loading hoses, or improper hose connections
These may generally be grouped into three categories for large facilities:
• Transfer, loading and unloading activities
• Processing activities
• Storage tanks and their spill control systems
Smaller facilities may not have any processing activities
Transfer areas include pipelines, pumps, valves, and control instrumentation needed to
achieve the movement of material within a facility. The loading/unloading area involves the
most handling operations and the largest potential for human error in most facilities. This is
where the raw materials are brought in and products and by-products are removed, and
temporary connections are frequently used. The storage area may be for raw materials,
intermediates, products, or by-products. The greatest volumes are contained here, so spill
sizes can be quite large. The processing area has equipment for raw material conversion into
products. This is the area that will only be found in a plant, while handling and storage
activities may take place at warehouses, water treatment facilities, greenhouses, and
numerous other types of miscellaneous facilities.
Examples of a broad spectrum of accidents are given below These cover events which
start with a release of hazardous materials, as well as those where a fire propagates into a
hazardous materials release.
• A natural gas pipeline rupture at a Texas refinery caused a series of
explosions and fires Two hundred firefighters worked nearly six hours to
control the blaze fed by three other propane pipelines Two refinery
workers were killed; a town of 1,200 people was evacuated. Damage to the
complex was predicted "into the millions " At least three similar explo-
sions had occurred in the area in the prior five years.
11-32
-------
Lightning struck a power pole in a chemical plant, then jumped to a tank of
dilute hydrochloric acid, damaging a tank valve. A cloud of gas floated
across several nearby neighborhoods. Hundreds of residents were forced to
evacuate their homes; three plant workers were injured (August 1,1980)
A refrigeration line at an ice cream plant in Burbank, California ruptured
spilling 300 gallons of anhydrous ammonia Firefighters sprayed water to
control the fumes until the main valve could be shut off and the leak
stopped. Eleven people were hospitalized and 60 residents were forced to
leave the surrounding area. (November 13,1984)
In Covington, Kentucky a chlorine gas cylinder ruptured at a swimming
pool jammed with about 300 swimmers More than 140 people were
hospitalized; no serious injuries were reported. (June 21,1981)
At least 11 people were injured and 100 persons evacuated from a
one-square mile area East Los Angeles after 100 gallons of chlorine
overflowed from a storage tank. (March 1,1982)
A 25,000 ton storage tank in Portland, Oregon, discharged 3-5 tons of
anhydrous ammonia due to a valve malfunction. An area three miles
downwind was evacuated while response personnel used water fog to
knock down vapors and had the spill vacuumed. (February 5,1982)
Falling equipment sheared off a pipe leading to a tank of hydrofluoric acid.
There were 66 senous injuries and roughly 3000 people were evacuated
around the Texas City oil refinery Water fog was used to help control the
vapor cloud and the tank contents were transferred to adjacent rail cars.
(October 31,1987)
More than 16,000 south Chicago residents were evacuated from the vicinity
of a bulk storage and terminal when a silicon tetrachlonde storage tank
sprang a leak. The escaping liquid, on the order of 150,000 gallons, reacted
with the moisture in the air to form hydrogen chlonde, resulting hi a dense,
corrosive, choking plume that stretched 5-10 miles downwind at times It
required 8 days for authorities to stop the leak, neutralize the spillage, and
transfer remaining tank contents to other containers. Approximately 100
people were hospitalized during the incident. (April 26,1974)
11-33
-------
• A truck driver delivering sodium hydrosulfide to a Chicago leather
company was directed to the wrong valve and began unloading the sodium
hydrosulfide cargo into a tank of chromic acid. The reaction of the two
chemicals released a deadly gas, hydrogen sulfide. Over 170 persons
working inside the four-story tannery were overcome by the gas; a total of
eight died and twenty-nine were injured. (February 14,1978)
There is one last type of accident that ments special attention for fixed facilities,
whether they be small or large. This is the potential for external events to cause releases,
with earthquakes being of particular concern Any natural disaster can cause releases as well
as affect responses For instance, transportation and access to the facility may be restricted,
water lines for fire protection may be broken, and resources may not be adequate to cover all
situations simultaneously Within any one site, an earthquake may impair the integrity of
containment (e.g., dikes) and/or may cause multiple containers/tanks to fail, thereby
exceeding the capacity of dikes, curbs, or other types of containment. Jurisdictions
particularly prone to such natural events as major earthquakes and floods should consider a
more formal analysis of facility nsks, taking into account the presence or lack thereof of
appropriate protective measures for these threats.
Suggested Approach for Assessment of Accident Potential
Based upon the information presented in Appendix F, the approach suggested for
getting a handle on fixed facility accident scenarios is to consider three basic types of release
events for plants; one or two release scenarios for facilities such as water treatment plants,
laboratories and industrial facilities; and one release scenario for warehouses and other
facilities storing hazardous materials. It has been shown that very little specific historical
information exists upon which to base accident rates. Hence, the best general approach is to
look at equipment failure rates. The increasing use of physical barriers to limit spills,
drainage systems to channel spills, and venting and scrubbing systems to control releases all
help to render this simplified accident estimation procedure more meaningful.
For example, a large facility may be coarsely modelled as having storage operations,
loading/unloading operations, and processing operations. These can respectively be repre-
sented by storage tank failures and leaks, hose failures, and piping and process vessel
failures. The rates suggested for each of these are.
Storage tank - double walled lO^/tank-year
Storage tank - single walled KH/tank-year
Pressure vessels lOYvessel-year
Inplant piping 1-5 x 10-«/ft-year
11-34
-------
Loading hoses lOVoperatton or
lO2/hose-year
While these certainly do not cover all potential release scenarios, they do capture some of the
more likely ways to lose large volumes of material. The only piping of prune concern is that
of relatively large diameter and long segments. In other words, a 100-foot expanse of 8" pipe
should be counted during assessment of failure potential if it contains hazardous materials,
but not 10 or 20 foot sections between vessels. As shown in Table 11.8, the spill size is
generally taken to be a function of the specific release scenario.
For the middle category of industrial users, water treatment plants, laboratories, etc.,
the main focus should be on storage tank or container failures. Piping failures or loading
hose failures may be considered if there is a significant amount of piping (say over 100 feet)
or if there are frequent loading/unloading operations (say 10 or more per year). The rates to
be used are the same as those listed above and summarized in Table 11.8.
Storage of hazardous materials, such as in warehouses or greenhouses, may also result
in failures of storage containers, but the greater threat here is probably from a fire which
spreads to the storage area and results in release, ignition, explosion, and/or combustion of
stored materials (with attendant evolution of potentially toxic smoke). The occurrence rate of
such fires is suggested to be 10-Vyr resulting in a release of 10% to 100% of the stored
volume of hazardous materials -- as summarized in Table 118. This is one area in which
more specific local data and information would be particularly helpful for better definition of
scenarios and estimation of their likelihood. Worksheet 11.5 summarizes the overall
recommended procedure recommended for analysis of fixed facilities.
The data required for an analysis that generally focuses on the larger and/or more likely
(yet significant) events at fixed facilities are:
• Material(s) of concern
• Number, dimensions, capacities, and contents of storage containers or tanks
• Number, dimensions, capacities, and contents of other vessels with large
inventories of hazardous material (such as columns, separators, reaction
vessels)
• Size, length, and operating conditions of piping systems
• Number of unloading and loading operations per year, materials involved,
and transfer flowrates
11-35
-------
TABLE 11.8
SUGGESTED FIGURES FOR FIXED FACILITIES
ITEM
ACCIDENT RATE
SPILL SIZE
Chemical Plants
Double-walled storage tank
Single-walled storage tank or
pressure vessel
Piping
Loading hose
10 x 10«/tank-year
10 x lOVtank-year
100% of tune ~ total amount of typical contents
released instantaneously
90% of tune - release of contents through 1" hole
until leak can be plugged or otherwise terminated
10% of time — total contents released
instantaneously
15 x 10*/foot-year 90% of time - release through 1" hole in wall of
pipe until leak can be plugged or otherwise
terminated
10% of tune — complete rupture of pipe
lO'Vloading or unloading or 100% of time — release through full hose diameter
10-Vhose-year at loading/unloading rate until flow can be
terminated
Industrial Users, Laboratories, Water Treatment Plants
Storage tank/container
Piping (if more than 100 ft)
Loading hose (if used more than 10
times/yr)
lOxlOVtank-year
15 x lOVfoot-year
10^/operation or
102/hose-year
90% of time — as above for single-walled storage
tank or vessel
10% of time - as above for single-walled storage
tank or vessel
90% of time ~ release through 1" hole in wall of
pipe until leak can be plugged or otherwise
terminated
10% of time — complete rupture of pipe
100% of time - release through full hose diameter
at loading/unloading rate until flow can be
terminated
Warehouses and Other Storage Facilities
Storage containers (drams, cylinders,
etc)
10-3/year
90% of time ~ release of 10% of total stored
volume
10% of time - 100% loss of total stored volume
(i e, all containers combined)
11-36
-------
WORKSHEET 11.5
ESTIMATING FIXED FACILITY RELEASE FREQUENCIES
Hazardous Materials):
Number of Process Vessels/Single-Wall
Storage Tanks:
Number of Double-Walled Storage Tanks:
Length of Pipe:
Annual Number of Loadings/Unloadings:
(or number of hoses)
Spill Frequencies*
Process Vessels/Storage Tanks:
Double-Walled Storage Tanks:
Piping:
Loading/Unloading Hoses'
Spills by Size**
Process Vessels/Storage Tanks
10% of contents (1" hole):
100% of contents:
Piping
release through 1" hole:
release through full pipe diameter for
time needed for shutdown or until
associated tank is emptied:
Loading/Unloading Hoses
release through full hose diameter
at transfer rate for time needed for
shutdown:
Notes:
A=-
B=-
f^ _
Dl=-
D2=-
= Axl(H=
OR
H = D2xlO2 =
Ex 0.9=.
(ExO.lH
Gx0.9 =
Gx0.1 =
(feet)
_(spills/year)
_(spills/year)
_(spills/year)
_(spills/year)
(spills/year)
_(spills/year)
(spills/year)
_(spills/year)
_(spills/year)
H =
_(spills/year)
*Assumes that the consequences of releases will be based on the tanks, piping and loading
hoses which give the worst consequences If desired, individual components or smaller
groupings may be evaluated by completing the worksheet separately for each grouping.
+The user may consider all these scenarios for consequence modelling and planning
purposes, or just the largest spill in each category.
11-37
-------
The more of this data available, the better. A first cut analysis could start with just the
storage tanks/containers and any other large vessels, adding loading activities and piping
only if the more information is desired or needed. Tanks of similar size and contents may be
grouped together for probability analysis purposes, as may pipelines.
Additional Data/Methodologies
In addition to the multi-tiered approach discussed above, there are a number of more
detailed analysis procedures that are appropriate for fixed facilities. These generally require
more complete risk assessments which can be time consuming and expensive, and which
mandate the complete cooperation of plant/facility management in order to obtain accurate
and complete results. Numerous data sources discuss the merits and requirements of such
analyses. Some starting points might include CONCAWE: 1982, Atherton et al: 1980,
Simmons: 1974, COVO: 1982, and literally dozens of other reports and hundreds of articles.
Use of Results in Consequence Analysis
The choice of consequence procedures to be utilized for any given potential source of
hazardous material discharge or spillage depends on the type of container or tank involved
and the nature of the particular hazardous material contained therein. Although Chapter 12
of this guide and its related computer program provide a wide variety of analysis tools to
assist users in evaluating individual accident scenarios, the large number of diverse
situations that may be encountered at many fixed facilities does not permit the provision of
detailed guidance for each and every case. Rather, emergency planning personnel need to
apply the following guidelines with some degree of flexibility and common sense.
The first objective of a consequence analysis for a fixed facility should be evaluation of
threats posed by storage tanks or process vessels that contain significant amounts of
hazardous materials. Releases from short lengths (10-20 feet) of piping attached to these
containers are included in the vessel failure rates Each container of this type should be
evaluated individually. If a length of piping between containers can result in simultaneous
discharges from more than one container, this fact should be considered in the analysis. To
the extent possible, the analysis should utilize available information on the operating
conditions of each vessel in terms of typical and maximum amount of contents, operating
temperature, pressure, and so forth. Any container of a substance with the potential for
runaway polymerization, decomposition, or other unusual reactivity hazard should be given
special consideration and attention. (However, the frequency of such events cannot be
estimated using general historical data)
11-38
-------
Most intraplant or facility piping systems will be attached to one or more storage or
process vessels and should consider the amount of contents of such vessels. In addition,
larger lines of significant diameter and length may require analysis via use of pipeline
discharge models to the extent that is feasible, or alternatively, the use of tank discharge
models Although the procedures of Chapter 12 permit evaluation of full line breaks and
leaks from small holes in compressed gas pipelines, only full line breaks can be considered
for liquid pipelines Consequently, assistance will be required from facility owners or
operators for analysis of smaller liquid leaks if this lower consequence but higher probability
scenario is to be addressed. If shutdown capability is not known, a reasonable release
duration may be assumed to be 15 minutes
Loading hose failures commonly involve transfers to and from transportation vehicles.
It will be necessary in many (but not all) such cases to assume a discharge rate equivalent to
the cargo transfer rate through the hose, though the actual rate of discharge may be somewhat
higher (especially initially). The potential duration of discharge should be determined by
investigation of the presence of emergency shutdown systems, available response forces, and
the likely time needed for them to terminate the release by various means at their disposal. In
no case, however, should the total amount of discharge exceed the total contents of the
storage tank and the transportation vehicle. While evaluating such situations, pay special
attention to situations in which manually activated emergency shutdown systems may be
present but not approachable due to the hazardous properties of the material being released
Obviously, a truck driver may be reluctant or unable to approach a vehicle to pull an
emergency shutoff valve lever if his vehicle is venting large amounts of a highly toxic,
flammable, and/or explosive gas or liquid. In the absence of such information, a recom-
mended release duration is 5 minutes.
Evaluation of consequences from small containers or cylinders stored in warehouses,
laboratories, and a wide variety of other commercial or public facilities will require
consideration on a case by case basis where greater specificity is desired than that given in
Table 11.8 and Worksheet 11.5, giving due consideration to the nature of potential accidents
and their likely outcome. Gas cylinder leaks can generally be evaluated in a direct and
straightforward manner using the procedures of Chapter 12 and its related computer program
Note, however, that one of the most common major problems in this general type of facility
is not leakage from one or more containers but a fire in the facility that involves the
hazardous materials stored therein and produces large quantities of toxic smoke that may
adversely impact downwind areas. Given the diversity of substances that may be involved,
there is no simple manner in which the downwind area possibly requiring evacuation or other
protective action can be precisely estimated in such events. Rather the general emergency
plan must be sufficiently flexible to permit on scene decisions during an actual emergency
and the rapid initiation of evacuation or protective action activities.
11-39
-------
11.8 TRANSPORTATION OF PACKAGED HAZARDOUS MATERIALS
It has been estimated that there are up to 700,000 vehicles and vessels used to carry
hazardous materials hi small packages in the United States (OTA, March 1986). Such
packages may be cylinders, drums, barrels, cans, boxes, casks, bottles, or other similar
containers, and are defined as having a capacity of less than 110 gallons or 1000 pounds
They may be transported by air, water, rail (usually in box cars), truck or van Non-bulk
transportation of hazardous goods is estimated to represent 50% of the total tons shipped by
truck and 80% of the total truck spills (OTA, July 1986) Up to 8% of the marine
transportation of hazardous materials involves dry cargo barges carrying portable tanks or
drums.
Commodity flow information for such shipments is very limited due to the large
number and variety of shipments that take place and a lack of reporting requirements Thus,
there is also a lack of accident rates or spill amount distributions which can be applied on a
general basis. But the small amounts of materials involved usually (not always) imply
limited consequences in the event of a release.
Causes and Examples of Past Accidents
The causes of releases from small packages include:
• Improperly tightened or faulty fittings, valves, and closures
• Dropping packages (while loading/unloading or in transit)
• Puncturing packages (again while handling or in transit)
• Improper blocking and bracing which allows packages to move, fall or fail
from impact or crushing while in a vehicle
• Fires
• Freezing, getting wet, or other severe environmental exposure
Examples of some incidents involving small packages are given below. These
particular examples cover transportion as well as loading and unloading incidents
• Ten employees were treated and released when a container of arsenic
trichloride was found ruptured as a truck was being unloaded (July 19,
1985)
11-40
-------
• A drum loaded with pesticide began leaking its contents as it was being
transported on a flatbed truck near Cologne, Germany. The small spill
formed a large gas cloud which was toxic and corrosive. Traffic had to be
halted for a considerable time to allow clean-up crews to neutralize the
situation. (May 15,1985)
• A container loaded with drums of phosphorous was being lifted by crane
from a barge to the dock when it ignited after an adjacent obstruction
punctured a drum through the container walls. The crane driver submerged
the entire container in the water away from the barge. The phosphorous
reigmted when the container was placed on the dock; by then the fire
department had arrived and extinguished the fire with dry powder. No
deaths or injuries resulted. (May 15,1986)
• In a Houston port, a freight container explosion from improperly packaged
aluminum phosphide killed a dock worker. When inspectors from the
Marine Safety Office opened the container, boxes of high explosives were
found, even though the aluminum phosphide had warning labels marked,
"Do not store with high explosives." (January 20,1986)
• Amid heavy tourist traffic, a truck driver noticed that water was leaking
from his cargo, drums containing nuclear reactor sludge Further inspection
at a busy truck stop showed that the hole in the drum had been patched with
electrician's tape. (November 14,1979)
Suggested Approach for Assessment of Accident Potential
The available accident data covers accidents per unit of handling for a package
(without designation of spill or nonspill events) (Kloeber et al, 1979), and there are a few
data sources (ICF, 1984) which give release fractions for transit as well as at associated
terminals. The methods needed to take advantage of these data, however, require more
information than can be obtained by emergency planning personnel with a reasonable amount
of effort in the vast majority of cases
The approach suggested is to identify particular shipments of concern which were
discovered during initial data collection efforts and to only analyze those selected materi-
als/shipments which involve sufficient volumes of materials to pose major hazards The
basic approach is to then utilize the basic accident rates for rail, marine, or highway transport
as dictated by the type of transportation taking place, and to combine the resulting annual
accident rate with the fraction of accidents expected to result in a spill or discharge. As a first
pass on estimating the spill size distribution, consider using:
11-41
-------
• Loss of one container (or group of containers, such as 12 bottles in a box)
--90% of the time
• Full loss of cargo (all containers) — 10% of the time
Additional Data/Methodologies
Should more precise estimates be needed, one of the more detailed methodologies
should be utilized (for example, see ICF, 1984, and the references contained therein).
Use of Results in Consequence Analysis
Given the relatively small amounts of hazardous materials transported in individual
packages and the manner in which most fail, the best course of action for consequence
analysis purposes is to assume that the amount of cargo specified directly above is released
instantaneously to the environment
11.9 TRANSPORTATION OF HAZARDOUS MATERIALS BY AIR
The transportation of hazardous materials by air is generally limited to small packages,
but this category of accidents also applies to crop dusters applying pesticides. The materials
transported by air are usually of high value or of high priority time-wise. The annual tonnage
shipped is between 200 and 300 thousand tons, but this involves a very large number of
shipments One study (OTA, March 1986) reports that in an evaluation of the air cargo
packages at 39 major airports, 5% involved hazardous materials.
As demonstrated at the beginning of this chapter, there are relatively few hazardous
materials incidents each year involving this mode of transportation. One source (Hazardous
Materials Intelligence Report, May 31, 1985) found that there were only 6 incidents in 1984
which resulted in death, injury or more than $25,000 of property damage on a nationwide
basis. The accidents tend to concentrate in the vicinity of airports, as might be expected
There are numerous regulations covering the transportation of hazardous goods by air,
including quantity restrictions and detailed regulations involving packaging. Many incidents
which occur have been shown upon investigation to involve violations of these regulations
In terms of emergency response planning, there is little that can be done to accurately
determine a community's potential vulnerability to this type of accident, and admittedly,
there is a question of whether planning should go beyond the development of communica-
tions links with airport facilities that handle hazardous materials, and the identification of
11-42
-------
those locations where hazardous materials may be found on airport property This observa-
tion follows from both the very low frequency of serious incidents and the extremely
incomplete data available on commodity flows.
Crop dusters present a special case. Although accident rates and specific consequences
cannot be predicted in advance due again to a ranty of senous incidents, it is a good idea for
jurisdictions in which these aircraft work to give some thought to how to respond to a crash.
Note that these planes usually carry pesticides and herbicides that may be highly toxic when
concentrated. The crash site and surrounding areas may be highly contaminated. Crashes in
or near bodies of water may threaten water supplies or cause environmental damage.
11.10 SUMMARY
The preceding sections have presented methodologies for estimating the annual
probability of hazardous material releases from fixed facilities and transportation systems.
Individual worksheets have been provided for truck, rail, marine, pipeline and fixed facility
activities and operations. At the very least, a separate worksheet will need to be completed
for each material or group of materials of concern and/or each facility of concern. Even if
accident probabilities are not computed, the worksheets and related text can be valuable for
the identification and refinement of individual accident scenarios.
The most critical local information needed for completion of worksheets involves
exposure data - i.e., the number and length of shipments, the number and capacity of storage
tanks or process vessels, and so forth. When obtaining this information, it is usually
sufficient to obtain the correct order of magnitude - great expenditures of resources or very
precise counts are not warranted. Should information be unavailable, it may be necessary to
make a best estimate, with advice from other knowledgeable individuals. This is surely
better than eliminating any potentially significant accident scenario from consideration.
Another important point is that no one event can result in all possible potential
consequences simultaneously. In other words, if a rail car or tank truck expenences a
BLEVE, it cannot then also have major downwind toxic vapor dispersion hazards as these
require non-ignition of the cloud. Also, not all ignitions of flammable vapors result in
explosions. The percentage of events which result in various consequences is very dependent
on the material involved, the quantity released, and the reason for the release. General
percentage breakdowns cannot be given.
11.11 REFERENCES
Association of American Railroads. "Railroad Facts, 1985 Edition," August 1985.
11-43
-------
Atherton, J.G. et al "The Bulk Storage and Handling of Flammable Gases and Liquids,"
London: Oyez Publishing Limited, 1980
Chemical and Engineering News. "Emergency plans urged for railyard chemicals," July 29,
1985, p. 6.
Cheok, M.C., G.D. Kaiser and G.W. Parry. "Development of a Methodology for Compren-
sive Hazard Analysis - A Feasibility Study," prepared by NUS Corp. for the Federal
Emergency Management Agency, June 1985.
Clarke, R.K. et al. "Severities of Transportation Accidents," Sandia Laboratories, NTIS
SLA-74-0001, July 1976
CONCAWE. "Methodologies for Hazard Analysis and Risk Assessment in the Petroleum
Refining and Storage Industry," CONCAWE Report No. 10/82, Den Haag, December 1982.
Considine, M. "Risk Assessment of the Transportation of Hazardous Substances Through
Road Tunnels," Recent Advances in Hazardous Materials Transportation Research, An
International Exchange, State-of-the-Art Report 3, Transportation Research Board, Washing-
ton, DC, 1986, pp 178-185.
Considine, M., G.C. Gnnt and P L Holden. "Bulk Storage of LPG - Factors Affecting
Offsite Risk," Institution of Chemical Engineers Symposium Series No. 71, pp. 291-320.
COVO Steering Committee. Risk Analysis of Six Potentially Hazardous Industrial Objects
in the Rijnmond Area, a Pilot Study, Boston. D Reidel Publishing Co., 1982.
Federal Railroad Administration. "Accident/Incident Bulletin, No 156, Calendar Year 1987,"
July 1988.
Harvey, A E, P.C. Cordon and T S. Ghckman. "Statistical Trends in Railroad Hazardous
Materials Transportation Safety - 1978 to 1986," Publication R-640, Association of
American Railroads, Washington Systems Center, September 1987.
Hazardous Materials Intelligence Report, May 31,1985
ICF, Inc. "Assessing the Releases and Costs Associated with Truck Transport of Hazardous
Wastes," prepared for the U S Environmental Protection Agency, NTIS PB84-224468,1984
Kloeber, G. et al. "Risk Assessment of Air Versus Other Transportation Modes for
Explosives and Flammable Cryogenic Liquids, Volume I: Risk Assessment Method and
Results," prepared by ORI, Inc for Materials Transportation Bureau, NTIS PB80-138472,
December 1979.
11-44
-------
Lautkaski, R., T. Mankamo and M. Karkkainen. "Chlorine Transportation Risk Assessment,"
Revised Edition, Nuclear Engineering Laboratory, Report 27, Technical Research Centre of
Finland, September 1979.
Materials Transportation Bureau "Annual Report on Hazardous Materials Transportation,
Calendar Year 1983."
National Transportation Safety Board. "Railroad/Highway Grade Crossing Accidents Involv-
ing Trucks Transporting Bulk Hazardous Materials," NTIS PB82-113432, September 1981.
Nayak, P.R., D.B. Rosenfield and JH Hagopian. "Event Probabilities and Impact Zones for
Hazardous Materials Accidents on Railroads," prepared by Arthur D. Little, Inc. for the
Federal Railroad Administration, DOT/FRA/ORD-83/20, November 1983.
Office of Radiation Programs. "The Consequences and Frequency of Selected Man-Originat-
ed Accident Events," U.S. Environmental Protection Agency, NTIS PB80-211303, June
1980.
Office of Technology Assessment. "Transportation of Hazardous Materials," OTA-SET-340,
Washington, DC: U.S. Government Printing Office, July 1986
Office of Technology Assessment. "Transportation of Hazardous Materials' State and Local
Activities," OTA-SET-301, Washington, DC: U.S. Government Printing Office, March 1986.
Simmons, J.A. "Risk Assessment of Storage and Transport of Liquid Natural Gas and
LP-Gas," prepared by Science Applications, Inc. for the U.S. Environmental Protection
Agency, NTIS PB-247 415, November 1974.
Smith, R.N. and EL. Wilmot "Truck Accident and Fatality Rates Calculated from
California Highway Accident Statistics for 1980 and 1981," prepared by Sandia National
Laboratories for U.S. Department of Energy, SAND-82-7066, November 1982.
Sorensen, J.H. "Evacuations Due to Chemical Accidents: Expenence from 1980 to 1984,"
prepared by Oak Ridge National Laboratory, ORNL/TM-9882, January 1986.
Techmca. "Ethane and Ethylene Pipelines Between Mossmorran and Grangemouth, Assess-
ment of Residual Risk," Production No 9, London, January 1983
Transportation Systems Center. "Transportation Safety Information Report, 1984 Annual
Summary," U.S. Department of Transportation, DOT-TSC-RSPA-85-1, April 1985.
Urbanek, G.L. and E J. Barber "Development of Catena to Designate Routes for Transport-
ing Hazardous Materials," prepared by Peat, Marwick, Mitchell and Co for the Federal
Highway Administration, NTIS PB81-164725, September 1980.
11-45
-------
von Herberg, P. "Prevention is the Best Cure," Chemical Purchasing, September 1979, pp
79-84.
Wolfe, K E "An Examination of Risk Costs Associated with the Movement of Hazardous
Materials," submitted to the Transportation Research Forum's 26th Annual Proceedings,
October 22-24,1984.
Wright, CJ. and PJ. Student. "Understanding Railroad Tank Cars," Fire Command,
November 1985, pp. 18-21 and December 1985, pp. 36-41.
11-46
-------
12.0 CONSEQUENCE ANALYSIS PROCEDURES
12.1 INTRODUCTION TO ARCHIE
With the wide proliferation of personal computers throughout the United States in
recent years, particularly of IBM™ PC and fully compatible systems, it is now possible to
provide emergency preparedness personnel at all levels of government and industry with
relatively sophisticated computational tools to evaluate the nature and magnitude of threats
facing individual jurisdictions To facilitate what would otherwise be a difficult, time
consuming, and expensive (if not impossible) task in many cases, the majority of accident
hazard assessment and consequence analysis procedures required for a comprehensive hazard
analysis have been incorporated into a single software program titled Automated Resource
for Chemical Hazard Incident Evaluation (ARCHIE) A copy of Version 1.0 of the program
has been provided together with this guide Future efforts being considered by the federal
government to further refine ARCHIE include installation of a data base of chemical and
physical properties for a large number of hazardous materials, installation of more
sophisticated analysis procedures enabled by availability of a database, and development of a
version that will function on Apple™ Macintosh computers Whether or not these enhance-
ments are undertaken will depend upon the acceptance of ARCHIE by emergency
preparedness personnel and feedback received by the government on the usefulness of the
program to their individual planning efforts
-------
Due to the length of this chapter and the likelihood that users will refer to it often,
Table 12.1 provides a special index to the following sections
Purpose and Objectives of ARCHIE
The primary purpose of ARCHIE is to provide emergency preparedness personnel with
several integrated estimation methods that may be used to assess the vapor dispersion, fire,
and explosion impacts associated with episodic discharges of hazardous materials into the
terrestial (i.e., land) environment The program is also intended to facilitate a better
understanding of the nature and sequence of events that may follow an accident and their
resulting consequences
Be advised that the detailed site-specific modeling techniques incorporated within
ARCHIE differ from the more simplistic approaches in Technical Guidance for Hazard
Analysis and are likely to produce different results In addition, ARCHIE permits assess-
ments for numerous types of hazards not addressed in the earlier guide
General Features of the Program
The core of the ARCHIE computer program is a set of hazard assessment procedures
and models that can be sequentially utilized to evaluate consequences of potential discharges
of hazardous materials and thereby assist in the development of a basis for emergency
planning. In other words, ARCHIE can help emergency planning personnel understand 1)
the nature and magnitude of hazards posing a threat to their jurisdictions, 2) the sequence of
events that must take place for these threats to be realized, and ultimately 3) the nature of
response actions that may be necessary in the event of an emergency to mitigate adverse
impacts upon the public and its property Among the models or calculation procedures
incorporated into Version 1.0 of ARCHIE are
• Nine methods for estimating the discharge rate and duration of a gas or
liquid release from a tank or pipeline
• Seven methods to help the user estimate the size of any liquid pools that
may form on the ground
• Two methods to estimate the rate at which a liquid pool will evaporate or
boil and the duration of these phenomena until the point in time that the
pool is depleted.
• A method to estimate the size of the downwind hazard zone that may
require evacuation or other public protective action due to the release of a
toxic gas or vapor into the atmosphere
12-2
-------
TABLE 12.1
SPECIAL INDEX TO CHAPTER 12
121
122
123
12.4
125
126
127
128
129
1210
1211
1212
1213
1214
1215
12.16
1217
12.18
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
Section Title
ntroduction to ARCHIE
nstallation of the ARCHIE Computer Program
General Notes on Responding to Questions from the Program
Initialization of Program Configuration Settings
Display of the Program Title Screen
'ntroduction to Options on the Main Task Selection Menu
Introduction to the Hazard Assessment Model Selection Menu
Discharge Menu Option A' Non-Pressurized Rectangular Tank of Liquid
Discharge Menu Option B Non-Pressurized Spherical Tank of Liquid
Discharge Menu Option C Non-Pressurized Vertical Cylinder of Liquid
Discharge Menu Option D- Non-Pressurized Horizontal Cylinder of Liquid
Discharge Menu Option E. Pressurized Liquid When Discharge Location is 4
inches or Less from the Tank Surface
Discharge Menu Option F Pressurized Liquid When Discharge Location is
More Than 4 Inches from the Tank Surface
Discharge Menu Option G. Pressurized Gas Release from Any Container
Discharge Menu Option H. Release from a Pressurized Liquid Pipeline
Discharge Menu I* Release from a Pressurized Gas Pipeline
Hazard Model Menu Option B: Pool Area Estimation Methods
Hazard Model Menu Option C. Pool Evaporation Rate and Duration Esti-
mates
Hazard Model Menu Option D Toxic Vapor Dispersion Model
Hazard Model Menu Option E: Liquid Pool Fire Model
Hazard Model Menu Option F Flame Jet Model
Hazard Model Menu Option G Fireball Thermal Radiation Model
Hazard Model Menu Option H. Vapor Cloud or Plume Fire Model
Hazard Model Menu Option I Unconfined Vapor Cloud Explosion Model
Hazard Model Menu Option J Tank Overpressunzation Explosion Model
Hazard Model Menu Option K- Condensed-Phase Explosion Model
Remaining Options on the Hazard Assessment Model Selection Menu
Use of the Vapor Pressure Input Assistance Subprogram
Use of the Tank and Container Contents Characterization Subprogram
Other Computer Programs
Page
12-1
12-6
12-7
12-9
12-10
12-10
12-14
12-20
12-23
12-26
12-29
12-32
12-36
12-40
12-44
12-49
12-52
12-56
12-59
12-63
12-64
12-67
12-68
12-72
12-75
12-78
12-80
12-81
12-84
12-85
12-3
-------
• A method to evaluate the thermal radiation hazards resulting from the
ignition of a flammable or combustible pool of liquid
• Two methods to evaluate the size of the downwind area that may be
subjected to flammable or explosive concentrations of gases or vapors in air
due to the release of a flammable or explosive gas or vapor -- together with
the maximum weight of potentially explosive gas or vapor in air that occurs
during the incident.
• A method to evaluate the consequences of an unconfined vapor cloud
explosion if the flammable gas or vapor in air should explode upon
ignition.
• A method to evaluate the consequences of an explosion arising from the
internal overpressunzation of a sealed or inadequately vented tank due to
external heating or internal reaction
• A method to evaluate the consequences of an explosion arising from
ignition of a true explosive material in the solid or liquid state.
An overview of technical details for individual models and the assumptions applied during
their development are supplied in Appendix B for those who wish to review these aspects of
the models This chapter provides generalized discussions more suitable to the average user
Given a potential accident scenario defined during the Hazard Identification and/or
Probability Analysis portions of the overall hazard analysis, the program user is expected to
select the appropriate sequence of calculation procedures to be utilized (generally starting
from the top of the above list and working down) At the conclusion of the process, when the
user is satisfied that the scenario has been properly represented, the user may then ask for a
printed summary of the accident scenario evaluation results for future reference Subsequent
sections of this chapter will provide greater details on these topics
To facilitate conduct of accident scenario evaluations and organization or results,
assessment of each new scenario begins with the creation of what is referred to as an
Accident Scenario File (ASF). This is a computer data file that automatically stores both the
input data provided by the user and the results of all computations For convenience, the user
may assign any name to the fUe so long as the name is no longer than eight alphanumeric
characters. Files are automatically assigned the name extension " ASF" to differentiate them
from others on the computer Each is intended to consider a single specific accident scenario
involving a specific hazardous material The files are stored and retnved from the disk drive
and directory specified by the user during the system initialization step described below
12-4
-------
Please note that users should never rename an ASF file using DOS commands outside the
ARCHIE operating environment. Any such attempt will result in the file becoming
unusable until its original name is restored.
Once an ASF file has been created, it may be recalled, revised, and placed back in
storage under its original name, or it may be copied to an ASF file with a new name This
provides a great deal of flexibility to the ARCHIE program because.
• Users that are unsure of the validity of any particular input parameter value
during initial evaluation of a scenario may provide an estimate, determine
the proper value at their convenience, recall and correct the value in the
ASF file without having to start from scratch, and rerun the various models
as necessary to finalize the overall evaluation
• Users who wish to evaluate a series of scenarios that differ only slightly
from one another may recall the first file created, rename it, change the
input values associated with scenario differences, and rerun the appropnate
models with the new values. This procedure automatically creates a new
ASF file with minimal effort while leaving the original file unchanged
Accuracy and Limitations of Hazard Evaluation Methodologies
Several comments are in order with respect to the accuracy and other features of the
consequence modelling and hazard assessment procedures found in ARCHIE These
procedures are in many cases simplified versions of more sophisticated methods developed
by and/or available to professionals in the field. ARCHIE is intended to provide approximate
answers for general emergency planning purposes It will in most cases produce results that
overestimate rather than underestimate threats to a community, but occasional exceptions are
both possible and likely Application of safety factors by users is both encouraged and
recommended.
Although ARCHIE has the ability to address a wide variety of common accident
scenarios in a fairly comprehensive manner, it is not capable of addressing several
potentially hazardous phenomena that may result from accidents and which may therefore
require special consideration by emergency planning personnel. These limitations are
common to most if not all programs of this type and include inabilities to address:
• Downwind public exposures to toxic combustion products from fires.
• Damages to people or property from impact by high velocity fragments
produced in explosions.
12-5
-------
• Damages and injuries resulting from liquid superheat explosions or
significant explosions taking place inside a building or other structure
• Damages to property caused by exposure to thermal radiation or corrosive
substances.
• Unusual threats or phenomena associated with hazardous chemical reac-
tions
Be sure to carefully evaluate the accident scenario you select and ensure that it applies
to the hazardous material being studied
Use of ARCHIE for Mixtures of Hazardous Materials
The hazard evaluation methodologies in ARCHIE are primarily designed to address
spills or discharges of relatively pure substances Mixtures can only be handled by
knowledgeable users in certain special cases by provision of the physical and chemical
properties of the mixture where and when such properties are requested by the program
Inconsistent or Inappropriate Use of Models
A great deal of effort was devoted during development of ARCHIE to ensure that users
do not apply hazard evaluation methodologies in an inconsistent and/or inappropriate
fashion. Although anyone who attempts to use the program in such a fashion will find a
large number of checks and balances in the system to prevent misuse, the complexity of the
processes and phenomena being considered did not permit development of a fully foolproof
program It is therefore necessary for users to apply common sense at each stage of an
analysis to ensure that the input data and information provided to the program are reasonable
For example, it is obviously inappropriate to attempt to force the program to utilize one of
the fire or explosion models for a substance that is not inherently flammable, combustible, or
explosive. Postulation of accident scenarios that are beyond the realm of reasonable
credibility should be avoided to prevent unreasonable assessment results
122 INSTALLATION OF THE ARCHIE COMPUTER PROGRAM
Detailed instructions for installation of the computer program on a variety of IBM™ PC
and fully compatible personal computer systems are presented in Appendix E to this guide
Please refer to Appendix E and follow its instructions exactly prior to any attempt to run
the program. Although you are unlikely to harm the program diskette, it is also unlikely that
the program will operate as desired without reference to these instructions
12-6
-------
12.3 GENERAL NOTES ON RESPONDING TO QUESTIONS FROM THE PROGRAM
Throughout program use, the user will be required to answer numerous questions of
various types These typically involve
• Input of yes or no in response to simple questions
• Input of letters or numbers for selection of menu options
Input of numbers for measurable quantities
Response to Yes or No Questions
Whenever a question requires a simple yes or no answer, the query will generally end
with one of the following three phrases in parentheses Note that the key that must be
pressed after any entry to the program is variously referred to as the CARRIAGE RETURN,
RETURN, or ENTER key Although the program uses the abbreviation "" to represent
this key on screen displays, the following text will refer to this key as the ENTER key
• (Y or /N) ~ means that the program will accept either a "F1 or "y"
followed by a press of the ENTER key, or simply a press of the ENTER
key alone, to indicate a yes answer A "N" or "n" followed by a press of the
ENTER key is required to indicate a no answer
!*,„II
• (Y/N or ) — means that the program will only accept "Y" or "y
followed by a press of the ENTER key as meaning yes. A "N" or "n"
followed by a press of the ENTER key, or simply a press of the ENTER
key alone, indicates a no answer
• (Y/N) -- means that entry of an upper or lower case "y" or "n" followed by
a press of the ENTER key are the only acceptable responses
The choice of which of these options appears at the end of every question was made on
the basis of which answer is most likely to be provided by the user in any given situation
Thus, although the program asks the user numerous questions during an accident scenario
evaluation, a great many of them can be quickly answered by simply keeping the "little"
finger of the tight hand close to the ENTER key on the keyboard
Selection of Menu Options
All primary menus in ARCHIE present a list of options with each line item preceded by
a lower case letter. Selection of any specific option is accomplished by typing the
appropnate letter (either lower or upper case letters are acceptable) and following the entry
with a press of the ENTER key
12-7
-------
Several shorter menus displayed during use of various hazard assessment models
typically denote available options with a number instead of a letter. As in the pnor case,
selection of an option is accomplished by entry of the appropriate number followed by a
press of the ENTER key.
Entry of Required Input Parameter Data
During the course of an accident scenario evaluation, the user will be asked to provide
a variety of numerical input data about the hazardous material involved and the circum-
stances under which it may pose a hazard to the public. Each time that the computer program
requires a particular data item not previously provided by the program user, it will display a
custom tailored parameter input screen. These screens define and describe the data item, the
units in which the value is desired, and the range of values considered reasonable Each
screen also provides an opportunity for the user to confirm that the proper value has been
entered, and several will ask if the user desires assistance in estimating an appropriate value
A yes answer to one of these questions will activate one of several help sections of the
program. These ranges from one or two screens of text that provide guidance to major
subprograms that provide assistance in estimating the vapor pressures of hazardous materials
at various temperatures from minimal input data or which assist the user in characterizing the
tank or container in which the hazardous material resides and the physical states, weights,
and volumes of container contents.
Once a user provides an input value to a model at the request of the program, that value
is immediately stored in the ASF file. If the same value is required by a subsequent hazard
assessment model, the value is shown in a list of previously stored data (applicable to the
model being used) together with a query as to whether the user wishes to change any value
before continuing. It is VITALLY important to realize that changing one of these values
AFTER earlier use may invalidate results of a pnor model Consequently, if a parameter
value is changed, it may become necessary to return to and rerun any pnor models that used
this value to ensure that the entire sequence of computations is based upon a consistent set of
input data.
Of particular interest is that ARCHIE generally requires no more data about a
hazardous material than can usually be found on a well written and complete material
safety data sheet (MSDS). Nevertheless, users with access to more detailed knowledge of
material characteristics and properties are in several cases given the opportunity to improve
the overall accuracy of scenario evaluation results by providing more exact input data This
is especially true in those cases where ARCHIE suggests use of typical values for needed
data not found on a MSDS.
12-8
-------
12.4 INITIALIZATION OF PROGRAM CONFIGURATION SETTINGS
Once the program is properly installed, it can be started by simply typing ARCHIE
followed by a press of the ENTER key at the appropnate system prompt A few seconds will
be required while the computer reads the program into memory from its storage location.
Indeed, such waits of a several seconds in duration for the reading and writing of program
and data files from disk storage locations are common during use of the computer program
and are no cause for concern
The first time that the program is executed, it will begin with a series of two to four
questions The answers to these questions will tell the program about the video display
attached to the system and the locations where you wish to store and retrieve data from
Accident Scenario Files (ASF)
The first question is straightforward and simply asks if a color computer video monitor
is attached to the system Although there are programming techniques available to determine
whether the computer contains a video adapter card capable of displaying color, no such
method exists to determine if the card is actually attached to a color monitor (Note Some
of you with monochrome monitors may wish to experiment with telling the program you
have a color monitor, simply to learn if you prefer the screen displays generated to the more
simple screens that will otherwise appear)
The second question only appears if you answered yes to the first question and may be
a bit difficult to answer if you have limited knowledge of the computer system you are using
The question itself is "Is an EGA card installed and operating in the EGA mode?" EGA is
an abbreviation for Enhanced Graphics Adapter, a video display board option for computers
that permits higher resolution graphics than the more typical boards found in older and/or
less expensive systems The presence of an operational EGA requires special program
instructions to set the colors of screen borders Do not be concerned whatsoever if you do
not know the answer to the question Just answer no and continue The absolute worst that
can happen is that some screens will have a black border in place of the more colorful border
that would otherwise appear Indeed, the very fact that black borders appear consistently is a
good clue that you should change your answer at some point m the future (Note We tell
you below how to change your answers to any of these questions once the program is
running Nothing is cast in stone1)
The third question (which could be the second at times) requests the letter of the disk
drive on which you wish to store and retrieve Accident Scenario File data Most of you
should not have any problem deciding where you wish these files to be stored For the rest of
you, until you can obtain assistance for a more expenenced computer user, here is some
simple advice that will keep you out of trouble
12-9
-------
• If you are running the program from two floppy diskettes and there are two
floppy drive slots in the front of the computer, place a blank or semi-blank
but formatted diskette in the drive on the nght (if the drives are side by
side) or lower nght (if one drive is on top of the other), close the door of the
drive by twisting the lever, and then answer B to the question
• If you are running the program from two diskettes, but there is only one
floppy drive in the system, answer C for the time being. This will result in
storage of ASF files in the root directory of the hard drive or card installed
in your system.
• If the entire program fits on one diskette and you are using a floppy, just
answer A to the question for now
• If the program is installed on a fixed or hard drive within the computer, just
press ENTER
The last question only appears when you have installed the program on a fixed or hard
drive within the computer. If you do not understand the question, simply press ENTER to
continue and all will be well
The program will save the answers to these questions and use them automatically next
time you use the system As noted above, however, you can change your answers quite
easily. If you are within the program, choose Option E from the Mam Task Selection Menu
(see below) to restart the initialization procedure If you are outside the program, you can
force an automatic initialization the next time the program is used by erasing the file named
ARCHIE.INI on your program disk
12.5 DISPLAY OF THE PROGRAM TITLE SCREEN
The title screen for the program is simply that Give it a look and press ENTER to
continue. This screen will be the first to appear in systems that were initialized during prior
use of the program. Otherwise, it will follow the questions described above
12.6 INTRODUCTION TO OPTIONS ON THE MAIN TASK SELECTION MENU
The title screen is followed by the Main Task Selection Menu leproduced in Table
12 2, this being the place where you decide which major task you wish to accomplish There
are six options available lettered "a" to "f' These are discussed m the following, but not
quite in the order that you might expect, primarily because your needs the first time or two
you use the program will be different than your needs at later times.
12-10
-------
TABLE 12.2
MAIN TASK SELECTION MENU
MAIN
TASK SELECTION MENU
a Start assessment procedure for a new hazardous material accident scenario
b Recall and modify data for previously considered accident scenario
c Print summary of accident scenario hazard evaluation after completion
d Proceed to system descnption menu
e Reset system configuration settings
f Terminate session
ENTER LETTER OF SELECTED OPTION (a-f)
TABLE 12.3
SYSTEM DESCRIPTION AND USE INSTRUCTIONS
TASK SELECTION MENU
SYSTEM DESCRIPTION AND USE INSTRUCTIONS
TASK SELECTION MENU
a Program purpose and objectives
b Suggested sequence of model use
c Creation and use of accident scenario files
d Recall and modification of previously created accident scenario files
e Printout of accident scenario hazard evaluation summaries
f Sources of assistance
g Return to main menu
ENTER LETTER OF SELECTED OPTION (a-g)
12-11
-------
Option D: Proceed to System Description Menu
Selection of Option D results in the appearance of the System Description and Use
Instruction Task Selection Menu which is reproduced in Table 12 3 and which has seven
options lettered "a" to "g" This part of the program provides bnef descriptions of various
program features and objectives as well as other general information about the program It is
included primarily for users who ignore pnnted manuals or program instructions or who may
receive the program without a pnnted copy of this guide It can also help users who may not
have used the program for a time and simply need their memory to be refreshed Selection of
any option but the last will display one or more screens of information about the program and
then return you to this menu Selection of the last option will return you to the Main Task
Selection Menu.
Option E: Reset System Configuration Settings
Remember the questions discussed concerning initialization of the program7 This
option on the Main Task Selection Menu permits you to change any of your previous
answers by restarting the initialization process
Option F: Terminate Session
Simply stated, selection of this option from the Mam Task Selection Menu says
"goodbye" and ends the program.
Option A: Start Assessment for a New Hazardous Material Accident
Scenario
When Option A is chosen from the Mam Task Selection Menu, the program begins the
process of creating a new Accident Scenario File (ASF). As reported earlier, this is a data
file that automatically stores both the input data provided by the user as well as results of all
computations. Initial file creation steps entail answers by the user to several questions, these
involving:
• The name to be assigned to the ASF file (mandatory)
• The name of the hazardous material of concern (optional).
• The name, location, or address of the facility or transportation route where
the postulated accident may occur (optional).
• The geographical latitude of the potential accident location (optional).
• The geographical longitude of the potential accident location (optional)
12-12
-------
• The date on which the scenario was evaluated (optional).
• A one to three line textual description of the accident scenario being
evaluated (optional).
• An indication of whether or not the hazardous material of interest is
flammable or combustible (mandatory).
The program will prompt the user for necessary information at each step of the ASF
file initialization process and provide opportunities to change or modify user responses to the
above quenes. When this process is complete, the program will proceed to the Hazard
Assessment Model Selection Menu described further below.
Option B: Recall and Modify Data for Previously Considered Accident Scenario
Once an Accident Scenario File has been created via use of Option A on the menu, it
may be recalled, copied and renamed (this is optional), modified, and eventually placed back
into storage by selection of Option B on the menu
Prior to display of the Hazard Assessment Model Selection Menu, the program
provides opportunities to:
• View the names of all ASF files stored at the location specified during
program initialization.
• Specify the name of the file that is to be retrieved
• Indicate whether he or she wishes to copy the data in the named file to a
new file with a different name, and
• Review and revise the file initialization data provided when Option A was
used to create the file.
Option C: Print Summary of Accident Scenario Hazard Evaluation After Completion
Option C of the Main Task Selection Menu permits the user to obtain a printed
summary of the overall results of the accident scenario evaluation for any scenario previously
analyzed using ARCHIE. This printout will be two to seven pages in length depending on
which evaluation procedures were used and the number of tables that were generated.
The printouts are formatted using standard commands of the BASIC program language
and should therefore print without problem on a wide variety of computer printers. It cannot
be guaranteed, however, that all output devices will behave as desired, particularly in the
12-13
-------
case of laser printers. Many brands of these devices require individual special software
"drivers" that are normally only provided with specific types of commercial word processing
and graphics generation programs. In some cases, the presence of a "print spooler" may
cause problems during printing.
While on the topic of printed output, it is also well to note that a printout of most
individual program screens can often be obtained by pressing the PrtSc, Print Scrn, or similar
key on the keyboard. Some earlier keyboards, particularly those provided with the first
generation of personal computers, may require simultaneous pressing of the Shift and PrtSc
keys. As in the case of scenario evaluation summaries, be advised of the possibility that this
method may not work with some laser printers.
12.7 INTRODUCTION TO THE HAZARD ASSESSMENT MODEL SELECTION
MENU
The hazard assessment models incorporated into the computer program are made
available to the user as options on the Hazard Assessment Model Selection Menu reproduced
in Table 12.4. The menu has 14 options lettered "a" to "n"
The first major step in use of the computer program to evaluate a specific accident
scenario requires selection of appropriate models from the model selection menu To
facilitate this task, Figure 12 1 provides a logic diagram pertaining to the most common spill
hazards associated with episodic discharges of hazardous materials Figure 12 2 is a related
directory of the models available in ARCHIE to evaluate these hazards As noted on the
diagrams in this figure, letters in parentheses within individual blocks refer to options
available on the Hazard Assessment Model Selection Menu In all cases, it is best to start at
the top of one of the charts shown in the figure and work downwards towards the
conclusion of each threat scenario. Note that the model selection charts can be viewed from
within the computer program by selection of Option M from the subject menu
The second step in most accident scenario evaluations in which fairly detailed
information is available with regards to the tank, pipeline, or other container that may
discharge its contents will usually involve use of a discharge rate and duration estimation
model. Table 12 5 reproduces the separate menu for these models that will appear when
Option A is chosen from the Hazard Assessment Model Selection Menu
The first four options (a through d) on the Discharge Model Selection Menu are
intended for use when the temperature of a liquid chemical within a tank or other container is
at or below its boiling point temperature, with the choice of any particular option depending
on the general shape of the container Containers are assumed to be non-pressunzed because
the vapor pressure of their liquid contents will be at or below ambient atmospheric pressure
12-14
-------
TABLE 12.4
HAZARD ASSESSMENT MODEL SELECTION MENU
HAZARD ASSESSMENT MODEL SELECTION MENU
a
b
c
d
e
f
g
h
i
J
k
1
m
n
Estimate discharge rate of liquid or gas
Estimate area of liquid pool
Estimate vaporization rate of liquid pool
Evaluate toxic vapor dispersion hazards
Evaluate pool fire radiation hazards
Evaluate fireball radiation hazards
Evaluate flame jet hazards.
Evaluate vapor cloud/plume fire hazards
Evaluate vapor cloud explosion hazards
Evaluate tank overpressunzanon rupture hazard
Evaluate solid/liquid explosion hazard
Review model descnptions
Review model selection charts
Return to mam menu
ENTER LETTER OF SELECTED OPTION
(a-n)
12-15
-------
N5
HAZARDOUS MATERIAL
ACCIDENT
CLOSED
CONTAINER
IN
FIRE
EXPLOSIVE
MATERIALS
AT RISK
RELEASE
ON LAND
FORMS POOL
PRESSURIZED
GAS
TO MR
ABE
LIQtJID VSPORS
TOXIC?
SEE IIQDID CR &S
EEIEflSE CF OCN3EHTS
AS AHEBOEKBIE
IS Gas
FLSMMSBIE?
IS
GftS TCKIC>
EXPLOSICN
HAZAFD
ABE
CONTESTS
FTAMfiPIE
CONTAINER
CVEBEEESSQRIZKnCN
EXPLOSION HAZfiRD
IGNmCNOF
LIQUID
DOWNWIND
TOXIC GAS
HAZARD
DOWNWIND
TOXIC VAPOR
HAZARD
FIREBALL OR
BLEVE HAZARD
VAPOR CLOUD
FIRE HAZARD
IF IGNITED
JET
HAZARD IF
IGNITED
VAPOR CLOUD FIRE
HAZARD IF IGNITED
D06«NWIND TOXEC
VRPCRCRGAS
HAZARD
POOL FIRE HAZARD
UPON IGNITION
CF LIQUID
EXPLOSION HAZARD
IN SEWERS AND
CONFINED SPACES
VAPOR CLOUD EXPLOSION
HAZARD IF IGNITED
VAPOR CLOUD EXPLOSICN
HAZARD IF IGNITED
FIGURE 12.1
HP AfTTTV QPTT.T. HA7ARm
-------
FIGURE 12.2
MODEL SELECTION CHARTS
FIRE/EXPLOSION MODELS
ONLY APPLY TO FLAMMABLE
OR COMBUSTIBLE MATERIALS
UPON THEIR IGNITION.
TANK OR PIPELINE
DISCHARGES
(A)COMPUTE DISCHARGE
RATE AND DURATION
LETTERS IN () REFER
TO OPTIONS IN MAIN
MODEL SELECTION
MENU.
J_
(B)COMPUTE POOL AREA
FOR LIQUID SPILLS
1
FOR VAPOR/GAS DISCHARGE
DIRECT TO ATMOSPHERE
(E)EVALUATE POOL
FIRE HAZARDS
1
(c)COMPUTE POOL
EVAPORATION RATE
(G)EVALUATE
FLAME JET
HAZARDS
(H)EVALUATE VAPOR CLOUD FIRE HAZARD
(i)EVALUATE VAPOR CLOUD EXPLOSIONS
1
(D)EVALUATE DOWNWIND TOXIC GAS
OR VAPOR DISPERSION HAZARDS
CLOSED TANK ENGULFED
IN FIRE
J_
(F) EVALUATE FIREBALL
RADIATION HAZARD
UPON TANK RUPTURE
1
(j) EVALUATE TANK
OVERPRESSURIZATION
EXPLOSION HAZARD
FIRE/EXPLOSION MODELS
ONLY APPLY TO FLAMMABLE
OR COMBUSTIBLE MATERIALS
UPON THEIR IGNITION.
SOLID OR LIQUID
EXPLOSIVES
_L
(K) EVALUATE EXPLOSION/
DETONATION EFFECTS
EXPLOSION MODELS ABOVE
DO NOT CONSIDER HAZARD
OF AIRBORNE FRAGMENTS!
*** SEE GUIDE ***
LETTERS IN () REFER
TO OPTIONS IN MAIN
MODEL SELECTION
MENU.
12-17
-------
TABLE 12.5
DISCHARGE MODEL SELECTION MENU
DISCHARGE MODEL SELECTION MENU
NON-PRESSURIZED TANKS CONTAINING LIQUID
a Rectangular tank
b Spherical tank
c Vertl cylindrical tank
d Horzl cylindrical tank
PRESSURIZED TANKS CONTAINING GAS AND/OR LIQUID
e Liquid discharge from tank when hole/pipe end 4 inches or less from tank.
f Liquid discharge from tank when hole/pipe end more than 4 inches from tank.
g Gas discharge from any tank.
RELEASE FROM A LONG PIPELINE
h Pipeline containing liquid under pressure.
i Pipeline containing gas under pressure
j Return to main model selection
ENTER LETTER OF SELECTED OPTION (a-j):
12-18
-------
Options "e" and "f' pertain to tanks or other containers of compressed liquefied gases,
and are generally applicable to situations in which the temperature of the liquid is above its
normal boiling point Option "e" should be used when the location from which the liquid is
expected to exit is four inches or less from the internal wall surface of the tank Option "f" is
more appropriate when the discharge outlet may be more than four inches from the internal
wall, as may occur when a pipe directly connected to a container breaks or ruptures some
distance from the vessel Option "g" on the menu is intended for use when the tank or
container (excluding long pipelines) only stores a compressed gas
Options "h" or "i" on the menu should be used to evaluate discharges from long
pipelines. The first applies to lines solely containing some type of liquid The second
applies to lines solely containing compressed gases
The next nine sections of this chapter provide information on use of the nine discharge
models listed under Options A to I on the Discharge Model Selection Menu The tenth
option of this particular menu (this being Option J) will return the user to the Hazard
Assessment Model Selection Menu The nine sections pertaining to discharge models each
have titles that begin with the prefix "12 JfX Discharge Menu Option X "
The discharge model descriptions are followed by descriptions of the remaining model
options available from the Hazard Assessment Selection Menu Each of these sections has a
title that begins with the prefix "12 XX Hazard Model Menu Option X " Subsequent sections
discuss and describe non-model related options available from this menu, special subpro-
grams, and utilities available to assist the user in describing and defining input parameter
values necessary for model use. A final section of the chapter provides information about
related computer programs that have been developed under sponsorship of the federal
government
Be advised that each of the model descriptions was intentionally written to stand alone
to the maximum extent possible, thus facilitating future reference to these discussions. It is
for this reason that there is a considerable degree of redundancy within the sections that
follow
As a final note before individual model selection options are introduced and discussed,
emergency planning personnel should realize that the most common hazardous material
likely to be encountered is automative gasoline, yet specific properties of this material are
generally difficult to locate in the literature. Based upon an evaluation of the 24 most
common components of a typical fresh unleaded gasoline blend, key properties that should
be provided to ARCHIE in the absence of more precise data, include
12-19
-------
• Molecular weight = 90.9
• Normal boiling point =114 9°F
• Specific Gravity = 0.64 at 68°F
• Vapor pressure = 82 mm Hg at 0°F
• Vapor pressure = 343 mm Hg at 68°F
Vapor pressure = 595 mm Hg at 100°F
• Lower flammable limit = 1 4%
• Heat of combustion = 18,570 Btu/lb
12.8 DISCHARGE MENU OPTION A: NON-PRESSURIZED RECTANGULAR TANK
OF LIQUID
Purpose of Model
Intended for use with liquids having temperatures at or below their respective normal
boiling points, this model is used to estimate the duration and average rate of liquid discharge
from a punctured or otherwise leaking rectangular tank or container A container should be
classified as rectangular if it resembles a box of tissues or a shoebox in general shape
Required Input Data
The following input parameter values and information may be requested during use of
this model.
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Length of the rectangular tank or container (ft)
• Width of the rectangular tank or container (ft)
• Height of liquid in the container as measured from its bottom (ft)
• Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous matenals are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
12-20
-------
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day. Even when not higher in the container, the
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role)
Prior to asking the user to provide the weight of liquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided in Section
12 29 of this chapter Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place. In the case of the amount of
liquid m the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present. In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available.
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion; for example, in the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous A yes answer to this question will halt use of the model and result in the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute.
A no answer will result in normal continuation of model use.
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to: 1)
determine or estimate the area of the expected outlet in units of square inches, and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who require further guidance.
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge The input parameter screen for the data
item will provide guidance in selection of an appropnate value
12-21
-------
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable. The specific gravity given at a temperature of 68°F (20°C) on a typical material
safety data sheet (MSDS) will be of more than acceptable accuracy in most cases
Model Results and Usage
Results of the model include the average rate of liquid discharge in pounds/minute, the
duration of discharge in minutes, the total weight of contents discharged in pounds, and the
physical state of the discharged material (which will always be liquid when this particular
discharge model is used)
Results of the model are typically utilized by the program as input parameters to the
pool area estimation methods (Hazard Model Menu Option B) available from the Hazard
Assessment Model Selection Menu
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released is at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given at the end of the procedure, prior to the point in time that results are
stored in the ASF file, to replace the computed duration of discharge with a shorter time
This is one way in which users can adjust discharge model results to account for situations in
which the discharge outlet being considered is actually above the bottom portion of the tank
or container. An alternative approach would be to specify a height of liquid measured
upward from the location of the expected discharge outlet and not from the bottom of the
tank or container
Another major assumption is that there is an opening somewhere m the top portion of
the tank or container that permits entry of air to fill the volume previously taken by liquid
discharged to the enviroment This assumption will be valid for any container that has some
sort of pressure equalization system to maintain standard atmospheric pressure above the
liquid surface while liquid is being pumped m or drawn out of the container In cases where
no such system exists or is operational, it is well to recognize that the model will estimate a
much shorter discharge time duration and much higher average discharge rate than would be
expected in the real world The reason for this is that the flow of liquid will be periodically
interrupted as air enters through the discharge outlet to fill the new vapor space created above
the liquid surface. The situation will m many respects resemble that which occurs when a
bottle or can of a soft drink is turned upside down and the liquid exits in a senes of spurts
rather than in a continuous and smooth flow pattern
12-22
-------
A final and relatively minor assumption is that the discharge outlet, be it a hole in the
side of the tank or a broken pipe attached to the container, is relatively close to the container
This assumption can lead to underprediction of discharge durations and overprediction of
discharge rates in cases when the discharge outlet is at the end of a complicated and/or
lengthy piping system attached to the container since such piping systems produce friction
that can slightly slow the flow of liquid and hence reduce the discharge rate
12.9 DISCHARGE MENU OPTION B: NON-PRESSURIZED SPHERICAL TANK OF
LIQUID
Purpose of Model
Intended for use with liquids stored at temperatures at or below their respective normal
boiling points, this model is used to estimate the duration and average rate of liquid discharge
from a punctured or otherwise leaking spherical tank or container resembling a ball of some
kind in general shape.
Required Input Data
The following input parameter values and information may be requested during use of
this model
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the tank or container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the spherical tank or container (ft)
• Height of liquid in the container as measured from its bottom (ft)
• Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day Even when not higher in the container, the
12-23
-------
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role).
Prior to asking the user to provide the weight of liquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents. Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided in Section
12.29 of this chapter. Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended.
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place In the case of the amount of
liquid in the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available.
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion, for example, in the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous. A yes answer to this question will halt use of the model and result in the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute.
A no answer will result in normal continuation of model use
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to* 1)
determine or estimate the area of the expected outlet in units of square inches, and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge The input parameter screen for the data
item will provide guidance in selection of an appropriate value.
12-24
-------
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable The specific gravity given at a temperature of 68°F (20°C) on a typical material
safety data sheet (MSDS) will be of more than acceptable accuracy in most cases
Model Results and Usage
Results of the model include the average rate of liquid discharge in pounds/minute, the
duration of discharge in minutes, the total weight of contents discharged in pounds, and the
physical state of the discharged material (which will always be liquid when this particular
discharge model is used).
Results of the model are typically utilized by the program as input parameters to the
pool area estimation methods (Hazard Model Menu Option B) available from the Hazard
Assessment Model Selection Menu.
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released is at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given at the end of the procedure, prior to the point in time that results are
stored in the ASF file, to replace the computed duration of discharge with a shorter time
This is one way in which users can adjust discharge model results to account for situations in
which the discharge outlet being considered is actually above the bottom portion of the tank
or container An alternative approach would be to specify a height of liquid measured
upward from the location of the expected discharge outlet and not from the bottom of the
tank or container
Another major assumption is that there is an opening somewhere in the top portion of
the tank or container that permits entry of air to fill the volume previously taken by liquid
discharged to the enviroment. This assumption will be valid for any container that has some
sort of pressure equalization system to maintain standard atmospheric pressure above the
liquid surface while liquid is being pumped in or drawn out of the container. In cases where
no such system exists or is operational, it is well to recognize that the model will estimate a
much shorter discharge time duration and much higher average discharge rate than would be
expected in the real world The reason for this is that the flow of liquid will be periodically
interrupted as air enters through the discharge outlet to fill the new vapor space created above
the liquid surface The situation will in many respects resemble that which occurs when a
bottle or can of a soft drink is turned upside down and the liquid exits m a senes of spurts
rather than in a continuous and smooth flow pattern
12-25
-------
A final and relatively minor assumption is that the discharge outlet, be it a hole in the
side of the tank or a broken pipe attached to the container, is relatively close to the container.
This assumption can lead to underproduction of discharge durations and overprediction of
discharge rates in cases when the discharge outlet is at the end of a complicated and/or
lengthy piping system attached to the container since such piping systems produce friction
that can slightly slow the flow of liquid and hence reduce the discharge rate
12.10 DISCHARGE MENU OPTION C: NON-PRESSURIZED VERTICAL CYLINDER
OF LIQUID
Purpose of Model
Intended for use with liquids stored at temperatures at or below their respective normal
boiling pouits, this model is used to estimate the duration and average rate of liquid discharge
from a punctured or otherwise leaking vertical cylindrical tank or container A tank or
container that would be classified as a vertical cylinder would resemble (in general shape) a
can of tuna fish or a can of soft drink sitting upright on a table Such tanks are the most
typical containers seen at facilities that store gasoline and/or fuel oils in above ground
vessels.
Required Input Data
The following input parameter values and information may be requested during use of
this model.
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the tank or container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the vertical cylindrical tank or container (ft)
• Height of liquid in the container as measured from its bottom (ft)
• Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year In selecting these temperatures, note that the temperature
12-26
-------
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day Even when not higher in the container, the
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role)
Prior to asking the user to provide the weight of liquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided in Section
12 29 of this chapter Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place In the case of the amount of
liquid in the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available.
There may be situations envisioned m which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion; for example, in the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid m the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous A yes answer to this question will halt use of the model and result in the
assumption that the entire contents of the container will be released to the environment m one
minute and that the average rate of discharge will be the weight of tank contents per minute
A no answer will result m normal continuation of model use
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to 1)
determine or estimate the area of the expected outlet in units of square inches, and 2)
compute the diameter of the equivalent circle having this area. Appendix A to this guide
provides assistance in this task for those who may require guidance
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge The input parameter screen for the data
item will provide guidance in selection of an appropriate value
12-27
-------
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern. Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable. The specific gravity given at a temperature of 68°F (20°C) on a typical material
safety data sheet (MSDS) will be of more than acceptable accuracy in most cases
Model Results and Usage
Results of the model include the average rate of liquid discharge in pounds/minute, the
duration of discharge in minutes, the total weight of contents discharged in pounds, and the
physical state of the discharged material (which will always be liquid when this particular
discharge model is used).
Results of the model are typically utilized by the program as input parameters to the
pool area estimation methods (Hazard Model Menu Option B) available from the Hazard
Assessment Model Selection Menu
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released is at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given at the end of the procedure, prior to the point in time that results are
stored in the ASF file, to replace the computed duration of discharge with a shorter time
This is one way in which users can adjust discharge model results to account for situations in
which the discharge outlet being considered is actually above the bottom portion of the tank
or container. An alternative approach would be to specify a height of liquid measured
upward from the location of the expected discharge outlet and not from the bottom of the
tank or container.
Another major assumption is that there is an opening somewhere in the top portion of
the tank or container that permits entry of an- to fill the volume previously taken by liquid
discharged to the environment This assumption will be valid for any container that has some
sort of pressure equalization system to maintain standard atmospheric pressure above the
liquid surface while liquid is being pumped in or drawn out of the container In cases where
no such system exists or is operational, it is well to recognize that the model will estimate a
much shorter discharge time duration and much higher average discharge rate than would be
expected in the real world The reason for this is that the flow of liquid will be periodically
interrupted as air enters through the discharge outlet to fill the new vapor space created above
the liquid surface The situation will in many respects resemble that which occurs when a
bottle or can of a soft drink is turned upside down and the liquid exits in a senes of spurts
rather than in a continuous and smooth flow pattern
12-28
-------
A final and relatively minor assumption is that the discharge outlet, be it a hole in the
side of the tank or a broken pipe attached to the container, is relatively close to the container.
This assumption can lead to underprediction of discharge durations and overproduction of
discharge rates in cases when the discharge outlet is at the end of a complicated and/or
lengthy piping system attached to the container since such piping systems produce friction
that can slightly slow the flow of liquid and hence reduce the discharge rate
12.11 DISCHARGE MENU OPTION D: NON-PRESSURIZED HORIZONTAL
CYLINDER OF LIQUID
Purpose of Model
Intended for use with liquids stored at temperatures at or below then: respective normal
boiling points, this model is used to estimate the duration and average rate of liquid discharge
from a punctured or otherwise leaking horizontal cylindrical tank or container. A container
classified as being a horizontal cylinder would resemble the tank seen on the back of a
typical gasoline truck in general shape.
Required Input Data
The following input parameter values and information may be requested during use of
this model.
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the tank or container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the horizontal cylindrical tank or container (ft)
• Length of the horizontal cylindrical tank or container (ft)
• Height of liquid in the container measured from its bottom (ft)
• Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
expenenced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
12-29
-------
ambient air temperature on a sunny day Even when not higher in the container, the
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role)
Prior to asking the user to provide the weight of liquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided in Section
12.29 of this chapter Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place In the case of the amount of
liquid in the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion, for example, in the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous. A yes answer to this question will halt use of the model and result in the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute
A no answer will result in normal continuation of model use
It should be recognized that not all horizontal cylindrical tanks have circular
cross-sections or flat ends as inherently assumed by the model Where the tank cross-section
is more of an oval, set the requested diameter equal to either the widest part of the tank (or to
the average of the narrowest and widest sections if somewhat greater accuracy is desired) It
is always safe to set the length of the tank to its longest dimension regardless whether its
ends are hemispherical, dished, or flat The errors introduced will be relatively minor.
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to 1)
12-30
-------
determine or estimate the area of the expected outlet in units of square inches; and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge The input parameter screen for the data
item will provide guidance in selection of an appropriate value
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable The specific gravity given at a temperature of 68°F (20°C) on a typical material
safety data sheet (MSDS) will be of more than acceptable accuracy in most cases
Model Results and Usage
Results of the model include the average rate of liquid discharge in pounds/minute, the
duration of discharge in minutes, the total weight of contents discharged in pounds, and the
physical state of the discharged material (which will always be liquid when this particular
discharge model is used)
Results of the model are typically utilized by the program as input parameters to the
pool area estimation methods (Hazard Model Menu Option B) available from the Hazard
Assessment Model Selection Menu
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released is at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given at the end of the procedure, pnor to the point in time that results are
stored in the ASF file, to replace the computed duration of discharge with a shorter time
This is one way in which users can adjust discharge model results to account for situations in
which the discharge outlet being considered is actually above the bottom portion of the tank
or container. An alternative approach would be to specify a height of liquid measured
upward from the location of the expected discharge outlet and not from the bottom of the
tank or container.
Another major assumption is that there is an opening somewhere in the top portion of
the tank or container that permits entry of air to fill the volume previously taken by liquid
discharged to the enviroment. This assumption will be valid for any container that has some
sort of pressure equalization system to maintain standard atmospheric pressure above the
12-31
-------
liquid surface while liquid is being pumped in or drawn out of the container In cases where
no such system exists or is operational, it is well to recognize that the model will estimate a
much shorter discharge time duration and much higher average discharge rate than would be
expected in the real world. The reason for this is that the flow of liquid will be periodically
interrupted as air enters through the discharge outlet to fill the new vapor space created above
the liquid surface. The situation will in many respects resemble that which occurs when a
bottle or can of a soft dnnk is turned upside down and the liquid exits in a senes of spurts
rather than in a continuous and smooth flow pattern.
A final and relatively minor assumption is that the discharge outlet, be it a hole in the
side of the tank or a broken pipe attached to the container, is relatively close to the container
This assumption can lead to underprediction of discharge durations and overprediction of
discharge rates in cases when the discharge outlet is at the end of a complicated and/or
lengthy piping system attached to the container since such piping systems produce faction
that can slightly slow the flow of liquid and hence reduce the discharge rate
12.12 DISCHARGE MENU OPTION E: PRESSURIZED LIQUID WHEN DIS-
CHARGE LOCATION IS 4 INCHES OR LESS FROM THE TANK SURFACE
Purpose of Model
Intended for use with liquids stored at temperatures above their respective normal
boiling points, this model estimates \h&peak rate of discharge and duration of discharge from
a punctured or otherwise leaking tank or container of what must be considered a compressed
liquefied gas under the specified conditions of storage. A special qualification is that the
model is only appropriate for use when the discharge outlet or hole is four inches or less from
the inner wall surface of the tank or container The discharge model described under Option
F below should be used if the discharge location is more than four inches distant
Required Input Data
Primary data requirements for use of the model include:
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the tank or container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the horizontal cylindrical tank or container (ft)
• Length of the horizontal cylindrical tank or container (ft)
• Height of liquid in the container measured from its bottom (ft)
12-32
-------
Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day
Prior to asking the user to provide the weight of liquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided m Section
12 29 of this chapter. Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended.
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place In the case of the amount of
liquid in the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available
The user may request assistance in determining the vapor pressure of the liquid at the
point in time that the program asks for this parameter value See Section 12 28 below for a
description of the Vapor Pressure Input Assistance Subprogram
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion, for example, m the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous A yes answer to this question will halt use of the model and result m the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute
A no answer will result in normal continuation of model use
12-33
-------
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to* 1)
determine or estimate the area of the expected outlet in units of square inches; and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge. The input parameter screen for the data
item will provide guidance in selection of an appropnate value
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
little with changes in temperature for many liquids for which this model is applicable The
specific gravity given at a temperature of 68°F (20°C) on a typical material safety data sheet
will be of acceptable accuracy in most cases.
Model Results and Usage
Results of the model include the peak rate of liquid discharge in pounds/minute, the
duration of discharge in minutes based on the peak rate, the total weight of contents
discharged in pounds, and an indication of the expected physical state of the discharged
material Depending upon the material, environmental, and normal boiling point tempera-
tures involved, the model may indicate that either an airborne mixture of gas and liquid
droplets (i.e, aerosols) or liquid is being discharged
In the case of liquid discharges from the container, results of the model are normally
utilized by the program as input parameters to the pool area estimation methods (Hazard
Model Menu Option B) available from the Hazard Assessment Model Selection Menu. In
the case of airborne gas and aerosol mixture discharges from the container, the results may be
utilized as necessary for input to the toxic vapor dispersion model (Hazard Model Menu
Option D) and/or the vapor cloud or plume fire hazard model (Hazard Model Menu Option
H) on the menu In addition, the duration of gaseous discharge may be utilized by the flame
jet model (Hazard Model Menu Option F)
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released in at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given to the user at the end of the procedure, prior to the point in time that
results are stored in the ASF file, to replace the computed duration of discharge with a
shorter time. This is one way in which users can adjust discharge model results to account
12-34
-------
for situations in which the discharge outlet being considered is actually above the bottom
portion of the tank or container. An alternative approach would be to specify a height of
liquid measured upward from the location of the expected discharge outlet and not from the
bottom of the tank or container. It is necessary to stress, however, that either of these
modifications to normal program use will result in a situation in which any gas exiting the
discharge outlet after completion of liquid discharge will not be accounted for in model
results This discharge of gas, which will quickly drop from a high to low rate when the tank
or container is not being heated or somehow internally generating heat, will pose a
downwind toxic or flammable gas hazard for a period of time that may be greater than that
estimated for the liquid discharge. If the remaining liquid in the tank or container is indeed
being heated, the flow of gas could be of considerable rate and duration yet incapable of
being estimated via use of ARCHIE It is for this reason that the model inherently assumes
that the entire tank will empty quickly and at a high rate to provide a conservative basis for
emergency planning purposes
A second assumption is that vaporization of liquid m the tank or container to generate
vapor or gas to fill the void left by escaping liquid will not result in a major thermodynamic
cooling effect. A significant cooling effect would tend to lower the estimated rate of
discharge and increase its duration, but this phenomena usually plays only a minor role in
influencing discharge rate and duration estimates The rate is primarily controlled by the
height of the liquid in the tank above the discharge outlet location and the specific gravity of
the substance
The decision as to whether the discharged material will be a mixture of gas and
aerosols or a liquid depends upon the temperature of the hazardous material in its container
and the normal boiling point of the substance. An airborne mixture of gas and aerosols is
assumed whenever the container content temperature exceeds the normal boiling point by
10 8°F (6°C) as a general rule of thumb The assumption is conservative in that it may at
times indicate that no liquid will reach the ground although this may indeed occur to some
degree More precise assessment of the physical characteristics of the discharged material
requires knowledge of physical property data not expected to be readily available to the
average user of Version 1.0 of ARCHIE and must therefore await installation of a database in
this program.
12-35
-------
12.13 DISCHARGE MENU OPTION F: PRESSURIZED LIQUID WHEN DIS-
CHARGE LOCATION IS MORE THAN 4 INCHES FROM THE TANK
SURFACE
Purpose of Model
Intended for use with liquids stored at temperatures above their respective normal
boiling points, this model estimates the peak rate of discharge and duration of discharge at
this rate from a punctured or otherwise leaking tank or container of a compressed liquefied
gas. A special qualification is that the model is only appropriate for use when the discharge
outlet or hole is more than four inches from the intarnal wall surface of the tank or container
The discharge model described under Option E should be used if the discharge location is
four inches or less distant from the wall
Required Input Data
Primary data requirements for use of the model include
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the tank or container (°F)
• Ambient environmental temperature (°F)
• Weight of liquid in the tank or container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the horizontal cylindrical tank or container (ft)
• Length of the horizontal cylindrical tank or container (ft)
• Height of liquid in the container measured from its bottom (ft)
• Diameter of the hole from which liquid will discharge (inches)
• Discharge coefficient of the hole
• Specific gravity of the liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day.
The user may request assistance in determining the vapor pressure of the liquid at the
point in time that the program asks for this parameter value See Section 12 28 below for a
description of the Vapor Pressure Input Assistance Subprogram
12-36
-------
Prior to asking the user to provide the weight of hquid in the tank, an input parameter
value that would otherwise require several manual computations, the program asks if the user
desires assistance in characterizing the volume of the container and the weights, volumes,
and physical states of its contents. Further details on the Tank and Container Contents
Characterization Subprogram and its general data requirements are provided in Section
12 29 of this chapter. Use of the subprogram may result in requests for a few informational
items not listed above but is nevertheless highly recommended
Hazard evaluations for emergency planning purposes should strive to assume the worse
credible conditions under which an accident may take place In the case of the amount of
liquid in the tank or container, guidance should be obtained from the owner or operator of the
vessel with respect to the largest amounts that may be expected to be present. In the case of
transportation vehicles, it is usually sufficient to assume that the container is at least 90% full
unless other information is available
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion; for example, in the event of its
exposure to a major explosion or collapse due to severe structural failure. Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous. A yes answer to this question will halt use of the model and result in the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute.
A no answer will result in normal continuation of model use
The program inherently assumes that the discharge outlet is circular in shape. In those
instances where the expected shape is not expected to be circular it will be necessary to. 1)
determine or estimate the area of the expected outlet in units of square inches, and 2)
compute the diameter of the equivalent circle having this area. Appendix A to this guide
provides assistance in this task for those who may require guidance.
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge. The input parameter screen for the data
item will provide guidance in selection of an appropriate value.
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for many liquids for which this model is applicable.
The specific gravity given at a temperature of 68°F (20°C) on a typical material safety data
sheet will be of acceptable accuracy in most cases.
12-37
-------
The specific heat capacity of the hquid is a measure of the heat necessary to raise the
temperature of a unit weight of the material by one degree Prior to request of this value, the
program will provide the user an opportunity to request assistance in selection of an
appropriate heat capacity value using several generalized rules-of-thumb (reproduced in
Table 12.6) that require user knowledge of the chemical formula of the hazardous material
being evaluated. Although the accuracy of model answers would be unproved by provision
of a more precise value for the substance at or near the temperature in its container, errors
introduced by use of an estimation method presented by the program should not be highly
significant in most (but not all) cases Note that liquid heat capacity values provided in units
of calories/gram-°C in various data bases and technical handbooks are numerically equiv-
alent to values expressed in units of Btu/lb-0F
Model Results and Usage
Results of the model include the peak rate of liquid discharge in pounds/minute, the
duration of discharge in minutes, the total weight of contents discharged in pounds, and an
indication of the expected physical state of the discharged material Depending upon the
material, environmental, and normal boiling point temperatures involved, the model may
indicate that either an airborne mixture of gas and liquid droplets (i e, aerosols) or a liquid is
being discharged.
In the case of liquid discharges from the container, results of the model are normally
utilized by the program as input parameters to the pool area estimation methods (Hazard
Model Menu Option B) available from the Hazard Assessment Model Selection Menu In
the case of airborne gas and aerosol discharges from the container, the results may be utilized
as necessary for input to the toxic vapor dispersion model (Hazard Model Menu Option D)
and/or the vapor cloud or plume fire hazard model (Hazard Model Menu Option H) on the
menu. In addition, the duration of gaseous discharge may be utilized by the flame jet model
(Hazard Model Menu Option F).
Major Assumptions of the Methodology
The model assumes that the hole or other discharge outlet from which the liquid is
being released is at or near the bottom of the tank for initial calculation purposes, thus
resulting in a further assumption of complete loss of liquid contents Note, however, that an
opportunity is given at the end of the procedure, pnor to the point m time that results are
stored in the ASF file, to replace the computed duration of discharge with a shorter time
This is one way in which users can adjust discharge model results to account for situations in
which the discharge outlet being considered is actually above the bottom portion of the tank
or container. An alternative approach would be to specify a height of liquid measured
upward from the location of the expected discharge outlet and not from the bottom of the
12-38
-------
TABLE 12.6
ASSISTANCE DISPLAY FOR LIQUID SPECIFIC HEAT
LIQUID SPECIFIC HEAT SELECTION ASSISTANCE
For organic materials predominantly consisting of carbon (C), hydrogen (H),
oxygen (O), nitrogen (N), and/or sulfur (S), values range from 0.30 to 0.80 A
value of 0.30 will usually give a conservative result and is suggested for use in the
absence of better data.
For materials containing chlonne (Cl), fluorine (F), or silicon (Si), values typically
range from 0 20 to 0.40 A value of 0 20 will usually give a conservative result
and is suggested for use in the absence of better data.
For materials containing bromine (Br) or iodine (I), or organic compounds
containing one or more metals such as nickel (Ni), iron (Fe), magnesium (Mg),
cadmium (Cd), tin (Sn), zinc (Zn), vanadium (Vd), or titanium (Ti), values
typically range from 0 10 to 0 20. A value of 0.10 will usually give an acceptable
result and is suggested for use in the absence of better data
TABLE 12.7
ASSISTANCE DISPLAY FOR VAPOR/GAS SPECD7IC HEAT RATIO
SPECIFIC HEAT RATIO SELECTION ASSISTANCE FOR GASES
Monatomic gases are substances such as argon, neon, xenon, krypton, or helium
Such gases have only one atom in each molecule when in the gaseous state. An
approximate specific heat ratio of 1.67 may be used for such gases when more
precise values are not readily available.
Diatomic gases are substances with two atoms in each molecule when in the
gaseous state. Examples include oxygen (0^, nitrogen (N^, hydrogen (Hj),
chlonne (Cy, and carbon monoxide (CO). An approximate specific heat ratio of
1.40 may be used when more precise values are not readily available for such
gases.
Polyatomic gases have more than two atoms in each molecule when in the gaseous
state Examples include ammonia (NlrQ, propane (CjjHg), and methyl bromide
(CHjBr) An approximate specific heat ratio of 1 30 may be used for such gases if
more precise values are not readily available
12-39
-------
tank or container. It is necessary to stress, however, that either of these modifications to
normal program use will result in a situation in which any gas exiting the discharge outlet
after completion of liquid discharge will not be accounted for in model results This
discharge of gas, which will quickly drop from a high to low rate when the tank or container
is not being heated or somehow internally generating heat, will pose a downwind toxic or
flammable gas hazard for a period of time that may be greater than that estimated for the
liquid discharge. If the remaining liquid in the tank or container is indeed being heated, the
flow of gas could be of considerable rate and duration yet incapable of being estimated via
this first version of ARCHIE It is for this reason that the model inherently assumes that the
entire tank will empty quickly and at a high rate to provide a conservative basis for
emergency planning purposes.
A second assumption is that vaporization of liquid in the tank or container to generate
vapor or gas to fill the void left by escaping liquid will not result in a major thermodynamic
cooling effect A significant cooling effect would tend to lower the estimated rate of
discharge and increase its duration, but this phenomena usually plays only a minor role in
influencing discharge rate and duration estimates The rate is primarily controlled by the
height of the liquid in the tank above the discharge outlet location and the specific gravity of
the substance.
The decision as to whether the discharged material will be a mixture of gas and
aerosols or a liquid depends upon the temperature of the hazardous material in its container
and the normal boiling point of the substance. An airborne mixture of gas and aerosols is
assumed whenever the container content temperature exceeds the normal boiling point by
10.8°F (6°C) as a general rule of thumb. The assumption is conservative in that it may at
times indicate that no liquid will reach the ground although this may indeed occur to some
degree. More precise assessment of the physical characteristics of the discharged material
requires knowledge of physical property data not expected to be readily available to the
average user of Version 10 of ARCHIE and must therefore await installation of a database in
this program.
12.14 DISCHARGE MENU OPTION G: PRESSURIZED GAS RELEASE FROM ANY
CONTAINER
Purpose of Model
The primary purpose of this model is to estimate the peak rate of discharge and
duration of discharge at this rate when a compressed gas is being released to the atmosphere
from a tank or other container that is not a long distance pipeline
12-40
-------
The model may be used in a very approximate fashion to evaluate gas or vapor
discharge rates and durations when a runaway exothermic chemical reaction of some kind
takes place in the container Use of the model for the latter purpose requires that the user
provide the program with suitably high values for the pressure and temperature in the tank
(accuracy in providing these values is not extremely important) and an appropriate molecular
weight and ratio of specific heats for the gas actually being discharged If the reaction
involves polymerization of one or more substances, some thought should be given to
reducing the weight of hazardous matenal(s) presumed to be in the tank by at least 10 to 33%
since the portion that polymerizes during the reaction is unlikely to be discharged to the
atmosphere as a gas or vapor.
Required Input Data
Primary data requirements for use of the model include
• Normal boiling point of the liquid (°F)
• Temperature of the gas and/or liquid in the container (°F)
• Ambient environmental temperature (°F)
• Pressure of the gas in the container (psia)
• Weight of hazardous material in the container (Ibs)
• Indication of whether or not an instantaneous spill is to assumed
• Diameter of the hole from which gas will discharge (inches)
* Discharge coefficient of the hole
• Ratio of specific heats (Cj/Cv) for the gas
• Height of liquid in the container measured from its bottom (ft)
• Molecular weight of the gas and/or liquid
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day
The user may request assistance in determining the pressure of the gas at the point in
time that the program asks for this parameter value See Section 1228 below for a
description of the Vapor Pressure Input Assistance Subprogram Note, however, that some
compressed gas containers may normally be at temperatures in excess of the critical
temperature of the gas. This is a temperature above which the substance cannot exist in a
12-41
-------
hquid state regardless of the pressure applied Consequently, the term vapor pressure no
longer has meaning when a material is at a temperature above its critical temperature, the
vapor pressure input assistance subprogram is no longer applicable, and it is necessary for the
user to specify any required pressures.
Prior to asking the user to provide the weight of hazardous material in the tank, an
input parameter value that would otherwise require several manual computations, the
program asks if the user desires assistance in characterizing the volume of the container and
the weights, volumes, and physical states of its contents Further details on the Tank and
Container Contents Characterization Subprogram and its general data requirements are
provided in Section 12.29 of this chapter. Use of the subprogram may result in requests for a
few informational items not listed above but is nevertheless highly recommended
Hazard evaluations for emergency planning purposes should strive to assume the worse
reasonable and credible conditions under which an accident may take place In the case of
the amount of liquid in the tank or container, guidance should be obtained from the owner or
operator of the vessel with respect to the largest amounts that may be expected to be present
In the case of transportation vehicles, it is usually sufficient to assume that the container is at
least 90% full unless other information is available
There may be situations envisioned in which the tank or container of interest may be
expected to fail in a very quick and catastrophic fashion, for example, m the event of its
exposure to a major explosion or collapse due to severe structural failure Consequently,
after the user has supplied the weight of liquid in the tank, the program asks if the user
wishes to assume that the spill or discharge should be assumed as being essentially
instantaneous. A yes answer to this question will halt use of the model and result m the
assumption that the entire contents of the container will be released to the environment in one
minute and that the average rate of discharge will be the weight of tank contents per minute.
A no answer will result in normal continuation of model use.
The program inherently assumes that the discharge outlet is circular m shape. In those
instances where the expected shape is not expected to be circular it will be necessary to* 1)
determine or estimate the area of the expected outlet m units of square inches, and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge. The input parameter screen for the data
item will provide guidance in selection of an appropriate value
12-42
-------
The ratio of specific heats for the compressed gas or vapor is a thermodynamic
property related to the amount of heat necessary to increase the temperature of a unit weight
of gas or vapor by one degree under specified conditions Prior to request of this value, the
program will provide the user an opportunity to request assistance in selection of an
appropnate specific heat ratio using several generalized rules-of-thumb (reproduced in Table
12 7) that require user knowledge of the chemical formula of the hazardous material being
evaluated Although the accuracy of model answers would be improved by provision of a
more precise value for the substance at or near the temperature in its container, errors
introduced by use of the property estimation methods provided by the program should not be
highly significant in most (but not all) cases
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide.
Model Results and Usage
Results of the model include the peak dicharge rate of compressed gas or vapor
discharge in pounds/minute, the duration of discharge in minutes at this peak rate, the total
weight of contents discharged in pounds, and the physical state of the discharged material
(which will always be gas when this particular discharge model is used) It is vital that the
user read and understand the assumptions made during formulation and use of this model
Results may be utilized as necessary for input to the toxic vapor dispersion model
(Hazard Model Menu Option D) and/or the vapor cloud or plume fire hazard model (Hazard
Model Menu Option H) on the menu In addition, the duration of gas discharge may be
utilized by the flame jet model (Hazard Model Menu Option F).
Major Assumptions of the Methodology
As a compressed gas or vapor is discharged from a tank or other container only
containing the material in its gaseous state, the loss of gas to the atmosphere together with
thermodynamic cooling effects associated with gas expansion will typically result m a rapid
decrease of the discharge rate with time. Since this model bases its results solely on the
initial peak discharge rate of the gas or vapor, it will consistently overestimate the discharge
rate and underestimate the duration of discharge The overestimation of discharge rate may
result in overestimation of downwind toxic or flammable gas hazard zones (which is not bad
for emergency planning purposes) and underestimation of hazard durations (which must be
kept clearly in mind) when Options D and H are respectively chosen from the Hazard
Assessment Model Selection Menu Similarly, the flame jet model available as Option G on
the menu may underestimate the time duration of the flame jet on occasion.
12-43
-------
When the tank contains a liquefied compressed gas and the gas discharge model is
being used to evaluate a discharge of gas or vapor from the head space of the tank (i.e., the
volume above the liquid surface), there is a excellent chance that the computed peak
discharge rate will vary even more significantly from the actual average discharge rate,
primarily because the model assumes that all of the tank or container contents, including the
liquid portion, will discharge at the computed peak gas discharge rate In actuality, there
may be cases when thermodynamic cooling effects will rapidly cool the liquid to a
temperature near its boiling point, at which tune flow from the tank will greatly decrease and
even possibly stop while the tank still contains a considerable amount of hazardous material
The assumptions described above can be restated in a different fashion by noting that
the model formulation inherently assumes that sufficient heat is entenng the tank or container
from the external environment, or is being internally generated due to an internal chemical
reaction of some kind, to offset any thermodynamic cooling effects and to maintain the gas
discharge rate near its peak value. Since this model may be used at times to evaluate
discharge rates and durations under these "worst case" conditions, and any number of
accident scenarios may involve the potential for exposure of the tank or container to an
external fire or an internal exothermic chemical reaction, the assumptions clearly have
advantages as well as disadvantages.
The primary reason why it was necessary to provide a model that may provide
inaccurate results at times (but results that tend more often than not to overestimate the
hazards of the release) is that a more ngorous and formal analysis procedure would
necessitate a substantially larger program and chemical physical property data not readily
available to the expected average user of Version 1.0 of ARCHIE.
12.15 DISCHARGE MENU OPTION H: RELEASE FROM A PRESSURIZED
LIQUID PIPELINE
Purpose of Model
This highly simplified model is intended to provide an estimate of the discharge rate
and duration expected when a long distance pipeline totally filled with liquid completely
ruptures at some point along its route. No distinction is made between liquids that are below
or above their respective normal boiling points Since the topography of the land surface
over the pipeline route will greatly influence model results, and since the topography is likely
to vary over the route, it may be necessary to use this model several times for any given
pipeline, each time assuming a different pipeline break location near sensitive environmental
areas or population zones
12-44
-------
Required Input Data
Data requirements of the model include the following input parameter values, some of
which are unusual The program makes a special attempt to assist the user in selecting a
value wherever an unusual item of information is required, but does expect the user to have a
rather complete understanding of pipeline, pipeline contents, and pipeline route characteris-
tics Note, however, that the relatively uninformed user should not hesitate to use the model
Input parameter values not known can be guessed at and modified at a later time. An
opportunity is always provided to discard the results of any analysis when it is completed
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in the pipeline (°F)
• Ambient environmental temperature (°F)
• Molecular weight of the liquid
• Specific gravity of the liquid
• Vapor pressure of the liquid in the pipeline (psia)
• Length of pipeline that will empty m the event of a rupture (ft)
• Diameter of the pipeline (inches or feet)
• Maximum height of the liquid column in the line that will empty (ft)
• Total pressure of the liquid in the pipeline (psia)
• Pipeline shutdown time if a discharge is detected (minutes)
• Indication of whether line breaks at one end or along route
• Pumping rate of liquid through the pipeline (Ibs/mmute)
• Discharge coefficient of the hole
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both pipeline and ambient environmental temperatures among the highest that may
be experienced during a typical year
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide.
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern. Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable The specific gravity given on a typical material safety data sheet (MSDS) will be
acceptable in most cases
12-45
-------
The user may request assistance in determining the vapor pressure of the liquid at the
point in tune that the program asks for this parameter value. See Section 12.28 below for a
description of the Vapor Pressure Input Assistance Subprogram
The question regarding the length of the pipeline that will empty in the event of a
rupture can be rather tricky when the line travels up and down various hills and valleys The
length that empties will not necessarily be the total length of the pipeline from its beginning
to its end.
Picture a pipeline that starts at the top of one hill, travels down to the floor of a valley,
and then travels up the side of another hill to its top If the pipeline breaks at the top of either
hill, and we assume there is no pressure in the line for the moment, then it can be understood
that little if any of the pipeline contents will escape to the environment Alternatively,
picture what would happen if the pipeline were to break on the floor of the valley In this
case, all contents of both resulting pipeline sections from the top of one hill to the top of the
second will empty onto the valley floor Thus, it is important to carefully consider the
topography of the land and specific potential accident sites when dealing with long pipelines.
A tracing of the route on a topographical map of the area can be invaluable in estimating
appropriate pipeline lengths. It is prudent, however, to first ask the pipeline owner or
operator if a discharge volume analysis of any kind was conducted prior to installation of the
line since these analyses are sometimes necessary for compliance with environmental impact
reporting requirements associated with the obtainment of permits from regulatory authorities
A question related to the desired length of pipeline asks the user for the maximum
height of the liquid column in the pipeline that will empty upon pipeline rupture This is the
maximum vertical height between the assumed point of discharge and the highest point
within the pipeline from which liquid is expected to empty In the second example presented
above, it would be the vertical height fiom the top of the highest hill to the floor of valley
where the pipeline rupture is expected A highly accurate estimate is not necessary
The next question essentially requests the user to provide the total pressure of the liquid
in the pipeline. Provision of an answer requires the user to realize that the force that moves
the liquid from one end of the pipeline to the other is the pressure mechanically provided by
various pumping stations along the route. This pressure can be considerable and far above
the simple vapor pressure of the liquid as it is pushed up various hills Although the program
uses this pressure in units of psia, the user is given five sets of units to choose among while
providing an answer to the question
Many pipeline systems have sensors and alarms that alert operators to the fact that the
pipeline has developed a major leak and permit them to shutdown pumping stations
manually, while others have systems that act automatically to shutdown pumps Quicker
12-46
-------
shutdowns obviously result in less loss of pipeline contents since the pumps will cease to
push the liquid through the line. Thus, even a rough estimate of how long it will require the
owner or operator of the line to detect and act upon a major pipeline break will help improve
the accuracy of analysis results. This information should be readily available from pipeline
owners or operators.
If a pipeline breaks or ruptures at one end, then the liquid will be released from one
open end of the line Conversely, if the line breaks along its travel path, there could
conceivably be two open ends of the line from which liquid will flow, thus increasing the
discharge rate and reducing the duration of discharge It follows that the program requires
some indication of whether the discharge will involve one or two ends of the pipeline.
During the elapsed time between the instant that the pipeline breaks or ruptures and the
pumps are shut down, there will be some penod of time in which liquid continues to spill
from the line due to continued pumping. Consequently, the program asks the user for the
normal flowrate of liquid through the line under normal operating conditions This also is an
item of information that should be readily available from the owner or operator of the line
Although the program will use this rate internally in units of pounds per minute, the user is
given a choice of five sets of units in which to provide an answer
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge The input parameter screen for the data
item will provide guidance in selection of an appropriate value The proper value for
pipeline breaks will usually be 0.62.
Model Results and Usage
In cases where the pumping shutdown time is specified as zero, the program will
provide the average discharge rate in pounds per minute, the duration of discharge in
minutes, the total weight of discharged material in pounds, and an indication of whether the
pipeline contents will be discharged as an airborne mixture of gas and small liquid droplets
(i e., aerosols) or as a liquid. When the pumping shutdown time is greater than zero, a brief
paragraph will describe the discharge rates and times associated with various time periods
and provide an overall average rate and duration of discharge for use by subsequent models
In the case of liquid discharges from the pipeline, results of the model are normally
utilized by the program as input parameters to the pool area estimation methods (Hazard
Model Menu Option B) available from the Hazard Assessment Model Selection Menu In
the case of gas and airborne aerosol discharges from the container, the results may be utilized
as necessary for input to the toxic vapor dispersion model (Hazard Model Menu Option D)
12-47
-------
and/or the vapor cloud or plume fire hazard model (Hazard Model Menu Option H) on the
menu. In addition, the duration of gaseous discharge may be utilized by the flame jet model
(Hazard Model Menu Option F).
Major Assumptions of the Methodology
As noted earlier, this is a highly simplified model for estimating discharge rates and
durations from liquid pipelines
The first simplifying assumption made, one that is not highly consequential, is that the
liquid experiences no friction as it flows through the pipeline. In reality, the walls of the line
will produce friction that will tend to slow the outward flow of liquid when a break occurs.
Inclusion of friction into the model, however, would not produce significantly decreased
discharge rates except under very unusual circumstances for the types of full line break
scenarios being considered
A major assumption that can indeed adversely affect the accuracy of results involves
the inherent assumption that the liquid in the pipeline does not contain a dissolved gas under
high pressure and is not itself a liquefied compressed gas Liquids with dissolved gases
under pressure (such as a crude oil pipeline containing significant amounts of dissolved
natural gas) will experience what is sometimes referred to as a "champagne" effect upon
rupture. In other words, expansion of the dissolved gases when the pressure is relieved may
cause a gushing forth of pipeline contents to the extent that the total amount of liquid
discharged will be greater than that normally predicted by the model Similarly, liquefied
compressed gases may erupt from the pipeline in a manner not fully considered by the model.
The decision as to whether the discharged material will be a mixture of gas and
aerosols or a liquid depends upon the temperature of the hazardous material in its container
and the normal boiling point of the substance. An airborne mixture of gas and aerosols is
assumed whenever the container content temperature exceeds the normal boiling point by
10 8°F (6°C) as a general rule of thumb. The assumption is conservative in that it may at
times indicate that no liquid will reach the ground although this may indeed occur to some
degree. More precise assessment of the physical characteristics of the discharged material
requires knowledge of physical property data not expected to be readily available to the
average user of Version 10 of ARCHIE and must therefore await installation of a database in
this program.
12-48
-------
12.16 DISCHARGE MENU OPTION I: RELEASE FROM A PRESSURIZED GAS
PIPELINE
Purpose of Model
This model is intended to provide an estimate of the discharge rate and duration
expected when a pressurized long distance pipeline totally filled with a gaseous substance
develops a leak or completely ruptures at some point along its route. The model cannot be
used for pipelines containing both liquid and gas.
Required Input Data
Data requirements of the model include the following input parameter values:
• Normal boiling point of the substance (°F)
• Temperature of the gas in the pipeline (°F)
• Ambient environmental temperature (°F)
Vapor pressure of the gas in the pipeline (psia)
• Molecular weight of the gas
• Length of the pipeline from beginning to end (ft)
• Diameter of the pipeline (niches or feet)
• Actual pressure of the gas in the pipeline where applicable (psia)
• Indication of whether small hole or full rupture occurs in line
• Diameter of the hole from which gas will discharge (inches)
• Discharge coefficient of the hole
Ratio of specific heats (Cj/Cv) for the gas
As discussed m Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both pipeline and ambient environmental temperatures among the highest that may
be experienced during a typical year
The user may request assistance in determining the vapor pressure of the gas at the
point in time that the program asks for this parameter value See Section 12 28 below for a
description of the Vapor Pressure Input Assistance Subprogram Note, however, that many
gas pipelines, especially those conveying fuel gases with low normal boiling points, typically
operate at temperatures in excess of the critical temperature of the gas. This is a temperature
at or above which the substance canno^ exist in a liquid state regardless of the pressure
applied and permits pipeline operators to compress the gas to high pressures without causing
hquefication Consequently, the term vapor pressure no longer has meaning when a material
12-49
-------
is at a temperature above its critical temperature, the vapor pressure input assistance
subprogram is no longer applicable, and it is necessary for the user to specify the pressure in
the pipeline using information obtained from the pipeline owner or operator For purposes of
checking whether or not the temperature of the gas in the pipeline is above or below the
critical temperature, without actually asking for this value, the program estimates the critical
temperature of the gas as being 1.8 times the normal boiling point temperature of the
substance expressed in degrees Kelvin. (Technically oriented users may remember that the
rather common rule-of-thumb being applied typically uses a multiplication factor of 1 65
instead of 1.8. The higher factor was chosen to prevent the program from erroneously
assuming the critical temperature has been exceeded when the actual temperature is
somewhat greater than the value estimated when a factor of 1 65 is applied)
In those cases in which the critical temperature of the gas has not been exceeded, there
is a possibility that the pipeline might be operated at a pressure below the vapor pressure of
the material being transported Although this is not likely, the user is given the opportunity
to specify the actual pressure in the pipeline. Be advised that input of a pressure that exceeds
the vapor pressure of the pipeline gas at its specified temperature will result in a warning
message to the user to the effect that the pipeline contains liquid Since the model is only
applicable when there is no liquid in the line, it will be necessary for the user to either revise
the pipeline pressure to an acceptable value (i.e., one at or below the specified vapor
pressure) or to choose the liquid pipeline model for use (Discharge Menu Option H) As
noted above, neither of the available pipeline models can cope with a situation in which the
pipeline contains both gas and liquid
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2.8 of the guide.
The program gives the user the option of specifying that. 1) a relatively small hole
should be assumed to occur in the pipeline, 2) the line should be assumed to break
completely at one end; or 3) the line should be assumed to break completely at some point
along its route of travel. The diameter of the discharge outlet hole will only be requested if
the first option is chosen.
The program inherently assumes that the discharge outlet is circular in shape In those
instances where the expected shape is not expected to be circular it will be necessary to. 1)
determine or estimate the area of the expected outlet m units of square inches, and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
12-50
-------
The discharge coefficient of the hole is a measure of its edge characteristics that has a
major and direct influence on the rate of discharge. The input parameter screen for the data
item will provide guidance in selection of an appropriate value The proper value will
usually be 0 62 for a leaking or ruptured pipeline
The ratio of specific heats for the compressed gas or vapor is a thermodynamic
property related to the amount of heat necessary to increase the temperature of a unit weight
of gas or vapor by one degree under specified conditions Prior to request of this value, the
program will provide the user an opportunity to request assistance in selection of an
appropriate specific heat ratio using several generalized rules-of-thumb (reproduced in Table
12 7) that require user knowledge of the chemical formula of the hazardous material being
evaluated Although the accuracy of model answers would be improved by provision of a
more precise value for the substance at or near the temperature in its container, errors
introduced by use of the property estimation methods provided by the program should not be
highly significant in most (but not all) cases
Model Results and Usage
Results of the model include the peak rate of gas discharge in pounds/minute (adjusted
in the case of full line break scenarios), the duration of discharge in minutes at the computed
rate of discharge, the total weight of contents discharged m pounds, and the physical state of
the discharged material (which will always be gas when this particular discharge model is
used).
To be realized is that the initial peak rate of gas discharge will drop in magnitude
extremely rapidly in any scenario involving a full break or rupture of the pipeline In order to
avoid gross overestimation of the discharge rate and underestimation of the discharge
duration in such cases, the model provides a rate of discharge that is 75% of the initial peak
discharge rate The impact of the adjustment should not be cause for concern since the
model will in all likelihood continue to overestimate the discharge rate and this is not
undesireable for emergency planning purposes
Results of the model may be utilized as necessary for input to the toxic vapor
dispersion model (Hazard Model Menu Option D) and/or the vapor cloud or plume fire
hazard model (Hazard Model Menu Option H) on the menu In addition, the duration of gas
discharge may be utilized by the flame jet model (Hazard Model Menu Option F)
Major Assumptions of the Methodology
The first simplifying assumption made, one that is not highly consequential, is that the
flow of gas is not hindered by friction In reality, the walls of the pipeline will produce some
degree of friction that will tend to slow the discharge rate.
12-51
-------
A second assumption is that the model assumes that thermodynamic cooling effects
will also not reduce the discharge rate of gas
22.17 HAZARD MODEL MENU OPTION B: POOL AREA ESTIMATION METHODS
Purpose of Methods
In the event of a discharge that results in formation of a pool of liquid on the ground, it
is necessary to obtain an estimate of the area of the pool This estimate is required by the
pool evaporation (Hazard Model Menu Option C) that estimates the rate of vapor evolution
to the atmosphere and by the liquid pool fire model (Hazard Model Menu Option E) that
estimates the height of the flame and surrounding hazard zones
Required Input Data
Use of the model requires the following information
• Molecular weight of the liquid
• Specific gravity of the liquid
• Discharge rate of the liquid from its container (Ibs/minute)
• Duration of liquid discharge (minutes)
• Normal boiling point of the liquid (°F)
• Temperature of the liquid in its container (°F)
• Ambient environmental temperature (°F)
• Wind velocity (mph)
• Vapor pressure of the liquid at ambient temperature (mm Hg)
• Area to which the liquid may be restricted if a secondary
containment system is present (ft2)
• Other input parameter values discussed below
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described hi Section 2 8 of the guide
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable. The specific gravity given on a typical material safety data sheet (MSDS) will be
acceptable in most cases.
12-52
-------
The hquid discharge rate and duration of discharge are usually computed via use of an
appropnate model from the Discharge Model Selection Menu. Any attempt to estimate pool
areas before these input parameters have been computed will result in a warning to the user
and a question as to whether he or she wishes to continue the analysis.
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day. Even when not higher in the container, the
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role).
Assistance in selection of a wind velocity that is consistent with the atmospheric
stability class specification that will be required by one or more subsequent models is
provided during model use (upon request of the user) by display of the class selection chart
(reproduced here in Table 12.8). The longest downwind hazard distances along the
centerhne of the wind direction are usually obtained when stability class F is specified, and
this is usually a good choice for general emergency planning purposes ~ as is a wind velocity
specification of 4.5 mph. However, be advised that there may be exceptions to this general
rule when an evaporating liquid pool is the source of hazardous vapor emissions Higher
wind velocities typically produce greater evaporation rates, yet are usually associated with
atmospheric stability classes other than F Thus, the actual worse case for toxic vapor
dispersion hazard assessments may involve a different combination than suggested above for
some materials. In the absence of more precise historical meterological data for the region of
concern, an atmospheric stability class of D together with a wind velocity of about 10 mph
can be assumed to evaluate threats under more typical atmospheric conditions. (Note. The
wind velocity and atmospheric stability class are not always used by these models but are
always requested. They will definitely be needed by subsequent models so this poses no
additional burden on the user.)
The user may request assistance in determining the vapor pressure of the liquid at the
point m time that the program asks for this parameter value. See Section 12 28 below for a
description of the Vapor Pressure Input Assistance Subprogram.
12-53
-------
TABLE 12.8
ATMOSPHERIC STABILITY CLASS SELECTION TABLE
A — Extremely Unstable Conditions
B - Moderately Unstable Conditions
C - Slightly Unstable Conditions
D - Neutral Conditions*
E -- Slightly Stable Conditions
F ~ Moderately Stable Conditions
Surface Wind
Speed, mph
Less than 4 5
45-67
67-112
134
Greater than 134
Daytime Conditions
Strength of sunlight
Strong
A
A-B
B
C
C
Moderate
A-B
B
B-C
C-D
D
Slight
B
C
C
D
D
Nighttime Conditions
Thin Overcast
> or = 4/8
Cloudiness**
-
E
D
D
D
< or = 3/8
Cloudness
-
F
E
D
D
S3
(J1
f*
*Apphcable to heavy overcast conditions day or night
**Degree of Cloudiness = Fraction of sky above horizon covered by clouds
-------
Many facilities install secondary containment systems in the form of dikes, curbs,
sumps, or pits designed to restrict and contain the spreading of discharged liquids in the event
of an accidental spill The program will ask if such a system is present If the answer is yes,
the program will next ask the user to provide an estimate of the area to which liquid
spreading will be restricted.
In the event that the program decides it best to assume that the liquid will boil upon its
release to the environment, it will directly compute the area of the boiling pool In all other
cases, the program will first estimate the maximum credible pool area for non-boiling but
potentially volatile liquid pools based on the information provided. It will then proceed to a
menu that provides the user with a variety of options for specifying or estimating the
expected pool area
Once the area of the evaporating or potentially boiling pool has been determined, it
next becomes necessary (only in the case of combustible or flammable liquids) to estimate
the area of the burning pool that will occur if the liquid is ignited This first involves a
question to the user as to whether he or she wishes to assume the liquid is ignited
immediately upon release or after the pool has attained its maximum size, with the latter
choice resulting in the larger and more hazardous fire. The program will then proceed to
estimate the area of the resulting burning pool.
Model Results and Usage
Use of the model provides two results The first is the approximate area of the
expected boiling or evaporating liquid pool The second, only provided in the case of
combustible or flammable liquids, is the estimated area of the burning pool
The boiling or evaporating pool area is typically utilized as an input parameter to the
liquid pool evaporation rate and duration estimation model (Hazard Model Menu Option C).
In addition, this area is used to adjust estimated initial evacuation zone widths resulting from
use of the toxic vapor dispersion model (Hazard Model Menu Option D). The estimated area
of the burning pool is normally utilized as an input parameter to" the liquid pool fire model
(Hazard Model Menu Option E)
Major Assumptions of the Methodologies
Estimation of pool areas resulting from liquid spills is one of the most difficult and
error prone aspects of accident scenario evaluations for hazardous materials, except in those
cases m which the discharge source is confined by a secondary containment system of known
dimensions and the liquid will cover the the bottom of the confinement area Unconfined
spills rarely occur in a location where the ground surface is flat and impermeable to liquids
Rather, in the real world, and particularly in transportation accidents, the spilled liquid will
12-55
-------
usually follow rainwater drainage paths while simultaneously vaporizing, burning, and/or
soaking into the ground. Thus, this model is actually comprised of a number of different
estimation procedures, all designed to ease the task of the program user in obtaining a
reasonable though not always highly accurate answer.
In the case of liquids that are expected to boil upon release to the environment, the
program uses a relatively simple and crude methodology to estimate the rate at which the
liquid will vaporize. It then uses this rate in conjunction with a liquid spreading model to
estimate the desired pool area. Various assumptions made are described in Appendix B.
For non-boiling liquids, the program first uses an evaporation rate model in conjunction
with a pool spreading model to estimate the pool area that would be necessary in order for
the total vaporization rate of the spilled liquid to approximate the discharge rate of the liquid
from its tank or other container. The result is a maximum credible pool area that provides an
upper bound that cannot be exceeded. The program then provides the user various options
for estimation of a more accurate pool area. Assumptions of these procedures are again of a
rather technical nature and require reference to Appendix B of this guide.
Burning pool areas are determined in a fashion similar to that used for boiling pools
with the exception that the vaporization rate of the pool is replaced with an estimate of the
rate at which the liquid will burn.
12.18 HAZARD MODEL MENU OPTION C: POOL EVAPORATION RATE AND
DURATION ESTIMATES
Purpose of Methods
Once the area of a boiling or evaporating pool of liquid has been estimated by the
above model or provided by the user, it next becomes necessary to obtain an estimate of the
total rate at which at which potentially hazardous vapors will evolve from the pool and the
duration of vapor evolution.
Required Input Data
Input parameter values that are most commonly requested by the program are listed
below. If the liquid is among a small group of substances with an extremely low normal
boiling point, only the boiling point, the discharge rate of the liquid m pounds per minute,
and the duration of discharge in minutes will be requested.
• Normal boiling point of the liquid (°F)
• Molecular weight of the liquid
• Specific gravity of the liquid
12-56
-------
• Temperature of the liquid in its container (°F)
• Ambient environmental temperature (°F)
• Vapor pressure of the liquid at ambient temperature (mm Hg)
• Area of the evaporating or boiling pool of liquid (ft2)
• Total weight of discharged liquid (Ibs)
• Atmospheric stability class (A to F)
• Wind velocity (mph)
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
very little with changes in temperature for the vast majority of liquids for which this model is
applicable. The specific gravity given on a typical material safety data sheet (MSDS) will be
acceptable in most cases
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day Even when not higher m the container, the
temperature of the liquid may increase upon spillage onto a hot surface or when exposed to
the sun (particularly when the vapor pressure of the material is relatively low and evaporative
cooling does not play a major role)
The user may request assistance in determining the vapor pressure of the liquid at the
point in time that the program asks for this parameter value See Section 12 28 below for a
description of the Vapor Pressure Input Assistance Subprogram
The area of the evaporating or boiling pool is usually computed via use of the pool area
estimation method listed as Option B on the Hazard Assessment Model Selection Menu
The total weight of spilled liquid is usually determined via use of one of the models available
from the Discharge Model Selection Menu Any attempt to use this model before these
parameters have been estimated by a prior model will result in a warning to the user and a
question as to whether he or she wishes to continue the analysis
12-57
-------
Assistance in selection of the appropriate atmospheric stability class and a wind
velocity consistent with this class is provided (upon request of the user) by reproduction of
the class selection chart (reproduced here in Table 12.8). The longest downwind hazard
distances along the centerhne of the wind direction are usually obtained when stability class
F is specified, and this is usually a good choice for general emergency planning purposes ~
as is a wind velocity specification of 4 5 mph However, be advised that there may be
exceptions to this general rule when an evaporating liquid pool is the source of hazardous
vapor emissions. Higher wind velocities typically produce greater evaporation rates, yet are
usually associated with atmospheric stability classes other than F Thus, the actual worst
case for toxic vapor dispersion hazard assessments may involve a different combination than
suggested above for some materials. In the absence of more precise historical meterological
data for the region of concern, an atmospheric stability class of D together with a wind
velocity of about 10 mph can be assumed to evaluate threats under more typical atmospheric
conditions. (Note: The wind velocity and atmospheric stability class are not always used by
this model but are always requested They will definitely be needed by subsequent models so
this poses no additional burden on the user)
Model Results and Usage
Results of the model include the rate of vapor evolution into the atmosphere in pounds
per minute and the duration of vapor evolution in minutes These results are typically
utilized as necessary as input parameter values to the toxic vapor dispersion model (Hazard
Model Menu Option D) and/or the vapor cloud or plume fire hazard model (Hazard Model
Menu Option H).
Major Assumptions of the Methodologies
As in the case of pool area estimation, estimation of pool vaporization rates and
durations from liquid spills is one of the most difficult and error prone aspects of accident
scenario evaluations for hazardous materials Although relatively sophisticated and accurate
estimation methods have been developed and validated in recent years, they unfortunately
require knowledge of chemical property data that are not always readily available and/or
entail complex algorithms that demand excessive computation times on typical personal
computers. Appendix B describes the rather unusual and relatively simple and crude
methodology employed by ARCHIE Be advised that answers will not be highly accurate
but will be in the correct "ballpark" for most cases Only occasionally should vaporization
rates be underestimated
12-58
-------
12.19 HAZARD MODEL MENU OPTION D: TOXIC VAPOR DISPERSION MODEL
Purpose of Model
The discharge, emission, or release of a toxic gas or vapor to the atmosphere has the
potential to pose an inhalation hazard to downwind populations under a wide variety of
circumstances The purpose of this model is to provide an estimate of the dimensions and
characteristics of the initial downwind zone that may require protective action in the event of
a hazardous material discharge.
Although primarily designed for gas or vapor releases, the model can be used in a very
approximate fashion for discharges of fine dusts or powders to the atmosphere. In such
cases, the user will be required to estimate and provide the program with the rate at which
these materials are emitted to the atmosphere and the duration of this emission
Required Input Data
The model requires eight input parameter values, these being.
• Molecular weight of the toxic substance
• Toxic vapor limit selected by the user (ppm, mg/m3, or gm/m3)
• Discharge height of the vapor or gas above groundlevel (ft)
• Atmospheric stability class (A to F)
• Wind velocity (mph)
• Temperature of the toxic substance in its container (°F)
• Ambient environmental temperature (°F)
• Vapor/gas emission rate (Ibs/rmnute)
• Duration of vapor/gas emission (minutes)
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide.
Selection of an appropriate toxic vapor limit is discussed in Chapter 6. Note that the
program will provide you with a choice of units in which this toxic limit may be expressed
All subsequent presentations of results, however, will utilize units of parts per million (ppm)
by volume in air.
Assistance in selection of the appropriate atmospheric stability class and a wind
velocity consistent with this class is provided (upon request of the user) by reproduction of
the class selection chart (reproduced here in Table 12 8) The longest downwind hazard
distances along the centerline of the wind direction are usually obtained when stability class
12-59
-------
F is specified, and this is usually a good choice for general emergency planning purposes --
as is a wind velocity specification of 4.5 mph However, be advised that there may be
exceptions to this general rule when an evaporating liquid pool is the source of hazardous
vapor emissions Higher wind velocities typically produce greater evaporation rates, yet are
usually associated with atmospheric stability classes other than F Thus, the actual worst
case for toxic vapor dispersion hazard assessments may involve a different combination than
suggested above for some materials. In the absence of more precise historical meterological
data for the region of concern, an atmospheric stability class of D together with a wind
velocity of about 10 mph can be assumed to evaluate threats under more typical atmospheric
conditions
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day.
The vapor or gas emission rate to the atmosphere and the duration of the emission are
usually computed by one or more of the preceding models on the Hazard Assessment Model
Selection Menu. Any attempt to use the toxic vapor dispersion model before these
parameters have been computed will result in a warning to the user and a question as to
whether he or she wishes to continue the analysis
Model Results and Usage
Estimation of hazard zone dimensions and characteristics requires a complex iterative
procedure that literally searches in the downwind direction for distances associated with the
user specified toxic limit concentration The program displays the results of this search as it
proceeds to find the downwind length of the hazard zone at the discharge height of the vapor
or gas, the peak concentration expected at groundlevel together with the location of this peak
for elevated sources, and the downwind length of the hazard zone at groundlevel. It then
provides a summary report of distances appropnate for the scenario being evaluated.
Two tables provide groundlevel concentrations, source height concentrations, recom-
mended initial evacuation zone widths, contaminant arrival times, and contaminant departure
times as a function of downwind distance.
12-60
-------
Due to the numerous uncertainties associated with even the most sophisticated vapor
dispersion modeling procedures, emergency planners should use the results of this model
only as guideposts in decisions regarding necessary protective action in the event of an actual
emergency. The pnme value of the model is hi its ability to provide a rough estimate of the
magnitude of downwind areas that will be at nsk, and in no way should model results be
expected to be either highly accurate or precise. Indeed, be advised that many professionals
consider the answers produced by a vapor dispersion model to be of acceptable accuracy if
they are correct within a factor of two in 50% or more of trials.
A special note is necessary with respect to the procedures used to estimate contaminant
arrival and departure times at downwind locations. Since the velocity of the wind generally
increases with height, and since the wind velocity reported by meteorologists is usually the
velocity measured at a height of 10 meters (about 33 feet) above the ground surface,
estimation of these times is at best highly approximate, particularly if the source of toxic gas
or vapor emissions is elevated above the ground The model will generally underestimate
rather than overestimate contaminant arrival times, but exceptions to this general rule are
distinctly possible. Similarly, the model will generally overestimate rather than underesti-
mate contaminant departure times, but exceptions to this general rule are again possible
under certain circumstances. See Appendix B for details of the computation procedures
Yet another special note is necessary with respect to the evacuation zone widths
predicted by the model when the wind velocities in the hazard zone are very low Under
such conditions, the direction of the wind can become very erratic, and it may not always be
wise to fully trust the results of the analysis in terms of its high probability that the wind will
not change direction within the first hour of the release. Be prepared at any time under low
wind conditions for one or more sudden shifts in wind direction and the possibility that the
cloud or plume of vapor or gas may literally "hop" from one position or direction to another.
Also, always remember that the evacuation zone widths are based on a probabilistic
evaluation of wind direction shifts Many a favorite race horse at the track has lost its event
by a wide margin. Longshots actually win at times.
Use of model results are rather self-evident. They provide emergency planning
personnel with an indication of the land area and thus populations subject to inhalation
exposures at or above the specified toxic limit concentration in the event of an accident In
addition, they provide an estimate of the duration of the toxic vapor or gas hazard These
items of information, in tarn, permit estimation of the number of people in the jurisdiction of
concern that may require notification, protective action, transportation, shelter, medical care,
and so forth.
12-61
-------
Mty'or Assumptions of the Methodology
The particular vapor dispersion model being utilized for toxic vapors primarily
assumes:
• The discharge is of finite duration.
• All vapors or gases are discharged from a single point
• The vapors or gases are neutrally buoyant.
• The ground surface is flat and generally free of obstacles.
• Atmospheric conditions remain constant during the discharge.
• The emission rate to the atmosphere is a constant during the release
• The gas or vapor enters the atmosphere at a low velocity
• Only gases or vapors are being released to the atmosphere
Because of the importance of this topic, Sections 3 5 and 3 6 of the guide are devoted
to a description and discussion of fundamental vapor dispersion phenomena and the factors
that influence the size and characteristics of downwind hazard zones Please refer to these
sections for an understanding of the significance of these assumptions. Also note that
Appendix B of the guide provides additional discussion of model assumptions, albeit m more
technical terms.
Be advised and stay aware that there is at least one special case in which the type of
dispersion model(s) being employed by ARCHIE may underestimate downwind hazard zone
lengths. This involves scenarios in which a compressed liquefied gas under very high
pressure in its container vents a high velocity jet of gas and liquid aerosols at a high rate m
the downwind direction. When the jet is forceful and the gas and aerosol mixture is
heavier-than-air, actual downwind travel distances to given concentrations m air may at times
exceed those predicted by ARCHIE
Due to the possibility that airborne contaminants released well above the ground
surface may be forced by local terrain effects to drop to groundlevel locations sooner than
would be expected, it is worthy to highlight the fact that evacuation zone width estimates are
based on the assumption that all discharges occur at groudlevel In other words, populations
residing under a cloud or plume of airborne contaminants are always assumed to be at risk
regardless of the fact that there may be times when the cloud or plume may harmlessly pass
over them.
12-62
-------
12.20 HAZARD MODEL MENU OPTION E: LIQUID POOL FIRE MODEL
Purpose of Model
The purpose of the hquid pool fire model is to compute the radius of the circular zone
around a fire in which unprotected and/or unsheltered people may expenence lethal burns
due to thermal radiation exposures. Additionally, the model computes the radius of the zone
in which second degree burns and/or severe pain may be experienced by exposed individuals
Required Input Data
The model requires four input parameter values, these being
• Molecular weight of the liquid
• Specific gravity of the liquid
• Normal boiling point temperature of the liquid (degrees F)
• Area of the burning pool (ft2)
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide.
The specific gravity of the liquid should ideally be the value associated with the liquid
at its temperature in the container or tank of concern Note, however, that this value vanes
little with changes m temperature for the vast majority of liquids for which this model is
applicable The specific gravity given on a typical material safety data sheet (MSDS) will be
acceptable in most cases.
The area of the burning pool is typically estimated using one of the pool area
estimation procedures available when Option B is selected from the Hazard Assessment
Model Selection Menu An attempt to use this model before this area has been computed
will result in a warning to the user and a question as to whether he or she wishes to continue
the analysis.
Model Results and Usage
The model provides the radius of the burning pool, the height of the expected flame,
the radius from the center of the pool in which exposed people may be fatality burned, and
the radius from the center of the pool in which exposed people may expenence second
degree burns or severe pain
12-63
-------
These results are useful to emergency planning personnel for estimating the number of
people that may require rescue and medical care, and for giving fire departments an
indication of the size and nature of the fire they must be prepared to confront. They are also
potentially useful for identifying other containers or stores of hazardous materials in the area
that may be subject to fire or thermal radiation exposures, possibly resulting in a tank
overpressurization explosion, a massive discharge of toxic gas, formation of a fireball, and/or
the high velocity dissemination of container fragments that may damage more distant
containers or populations.
Major Assumptions of the Methodology
The model assumes that the wind velocity in the burning pool area will be insufficient
to tilt the flame in the direction of the wind to a significant degree, thus resulting in circular
hazard zone estimates In the event the wind does indeed cause tilting of the flame, hazard
zones will be more of an oval shape and have a somewhat greater radius from the flame in
the downwind direction The radii of the respective zones will be somewhat smaller in the
upwind direction.
The model assumes that people in direct view of the flame and m the open will have
exposed skin. In other words, their skin will not be protected completely from the effects of
thermal radiation by any clothing being worn
Hazard zone estimates are based on the assumption that the base of the flame will be
fairly circular in shape. Any major deviation from such a shape may invalidate model
results, but these results are likely to remain conservative
The model assumes that neither carbon dioxide or water vapor in the air will absorb
any of the thermal radiation impinging on exposed people
Based on experimental data, a radiation intensity of 5 kW/m2 ( 1600 Btu/hr-ft2) was
selected for the purpose of defining injury zones since this incident flux will cause second
degree burn inj'unes on bare skin within 45 seconds An incident flux level of 10 kW/m2
(3200 Btu/hr-ft2) was chosen as the level capable of causing fatalities among exposed people
since it can be expected to quickly cause third degree burns leading to potential fatalities.
1221 HAZARD MODEL MENU OPTION F: FLAME JET MODEL
Purpose of Model
The flame jet model has the purpose of estimating the length of the flame and the zone
around it that may be subjected to harmful levels of thermal radiation when a flammable gas
discharges from its container at high speed and is ignited Due to special properties of the
12-64
-------
material, be advised that the model may greatly overpredict flame jet lengths associated with
discharges of hydrogen but is expected to produce relatively accurate estimates for other
substances.
Required Input Data
The model requires eight input parameter values, these being
• Normal boiling point of the material (°F)
• Temperature of the material in its container (°F)
• Ambient environmental temperature (°F)
• Pressure of the flammable gas in the container (psia)
• Ratio of specific heats (C,/CV) for the gas
• Molecular weight of the gas
• Lower flammable limit of the gas (volume %)
• Diameter of the hole from which the gas is venting (inches)
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day.
The user may request assistance in determining the pressure of the gas or vapor in the
container at the point in time that the program asks for this parameter value See Section
12 28 below for a description of the Vapor Pressure Input Assistance Subprogram
The ratio of specific heats for the compressed gas or vapor is a thermodynamic
property related to the amount of heat necessary to increase the temperature of a unit weight
of gas or vapor by one degree under specified conditions. Prior to request of this value, the
program will provide the user an opportunity to request assistance in selection of an
appropriate specific heat ratio using several generalized rules-of-thumb (reproduced in Table
12.7) that require user knowledge of the chemical formula of the hazardous material being
evaluated. Although the accuracy of model answers would be improved by provision of a
more precise value for the substance at or near the temperature in its container, errors
introduced by use of the property estimation methods provided by the program should not be
highly significant in most (but not all) cases
12-65
-------
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
described in Section 2 8 of the guide
The lower flammable limit (LFL) of a gas or vapor was defined in Chapter 4 of this
guide. Values are always listed on material safety data sheets (MSDS) when appropriate and
available for the material of concern
The program inherently assumes that the discharge outlet is circular m shape In those
instances where the expected shape is not expected to be circular it will be necessary to 1)
determine or estimate the area of the expected outlet in units of square inches, and 2)
compute the diameter of the equivalent circle having this area Appendix A to this guide
provides assistance in this task for those who may require guidance
Model Results and Usage
The model estimates the length of the expected flame jet and a safe separation distance
from the flame, both in units of feet If a gas discharge model providing a duration of
discharge has been used previously, the model will also display the expected duration of the
flame jet in minutes.
The length of the flame jet and its surrounding hazard zone radius are useful to
emergency planning personnel not only in estimating the number of people that may subject
to injury from the fire, but perhaps more importantly, the number and characteristics of other
containers or stores of hazardous materials at industrial sites that may be adversely impacted
by the fire. Of special concern is that a flame jet involving one storage or transportation
container may impinge on another container, thus possibly causing a tank overpressunzation
explosion, a massive discharge of toxic gas, formation of a fireball, and/or the high velocity
dissemination of container fragments that may damage more distant containers or popula-
tions
Major Assumptions of Methodology
The model assumes that the pressurized gas or vapor will discharge at sufficient
velocity to form a lengthy flame jet resembling a large torch In the event that this condition
is not satisfied, the flame exiting from the discharge outlet is likely to be shorter and to curve
upwards. In technical terms, this model is said to address turbulent or momemtum
dominated flame jets. If flow velocity criteria are not fulfilled, the flame will be buoyancy
dominated.
12-66
-------
Due to the complexities involved in estimating safe separation distances from flame
jets that may have a wide variety of orientations with respect to the person or persons at risk
of receiving burn injuries, the safe separation distance provided by the model is simply twice
the length of the flame This is considered to be a conservative estimate based on evaluations
conducted with more sophisticated models of flame jet phenomena
12 22 HAZARD MODEL MENU OPTION G: FIREBALL THERMAL RADIATION
MODEL
Purpose of Model
This model characterizes the fireball and associated thermal radiation hazard zones
resulting from exposure of a sealed or inadequately vented container of a flammable liquid or
liquefied compressed gas to an external fire or other source of excessive heat sufficient to
cause explosion or violent rupture of the container. The model may also be used in an
approximate fashion for containers of flammable compressed gas
Required Input Data
The model only requires the user to provide the weight in pounds of the flammable
material in the container at the point in time that it explodes or ruptures. The most
conservative course of action is to simply assume that the container will explode or rupture
when it is full. Realize, however, that many Boiling Liquid Expanding Vapor Explosions
(BLEVES) occur when flame weakens the wall of the vapor or head space of the container.
If the tank is fairly full to begin with and is fitted with a pressure relief device, it may vent a
considerable portion of its contents before occurrence of a BLEVE
Model Results and Usage
The fireball that results from rupture of a tank or container and subsequent ignition of
its contents will quickly nse while growing in size and then burn out. Results of the model
include estimates of the:
• Maximum diameter of the fireball (ft)
• Maximum expected height of the fireball (ft)
• Estimated duration of the fireball (seconds)
• Distance (radius) from the container hi which fatalities may be expected
due to thermal radiation burns (ft)
12-67
-------
• Distance (radius) from the container in which significant burn injuries may
be expected due to thermal radiation (ft)
Output of the model during the accident scenario evaluation session always stresses
that "Boiling Liquid Expanding Vapor Explosions (BLEVES) may cause high velocity tank
fragments to travel considerable distances Some tanks, especially horizontal cylindrical
types, may rocket while spewing forth flames" See Chapter 5 for a more complete
description and discussion of potential fragment hazards.
Results of the model are useful to emergency planning personnel in defining safe
separation zones in situations where there is a potential for fireball formation and fragment
impact hazards. Additionally, where the event may occur without time for protective actions
to be taken, the results can help planners estimate the number of people that may be killed or
injured. Although specific characteristics and potential damages cannot be predetermined,
the fact that container fragments may disperse at high velocity can help identify other
containers or stores of hazardous materials that may be impacted by the accident and pose
additional threats to the public.
Major Assumptions of the Methodology
Key assumptions of the model include:
• No thermal radiation will be absorbed by water vapor or carbon dioxide gas
present in the atmosphere
• All flammable materials of interest are similar in characteristics to lique-
fied compressed propane.
• Both the container and exposed people are on or near the ground
• The burn seventy depends upon the amout of energy absorbed by the skin
after a surface temperature of 55°C is achieved See Appendix B for
additional information.
1233 HAZARD MODEL MENU OPTION H: VAPOR CLOUD OR PLUME FIRE
MODEL
Purpose of Model
A plume or cloud of flammable vapor or gas has the potential to either burn or burn and
then explode upon encountering a suitable source of ignition. The purpose of this model is to
estimate the dimensions of the downwind area that may be subjected to flammable and
12-68
-------
potentially explosive vapors or gases in the event of an accidental discharge In addition, the
model estimates the maximum weight of flammable of explosive gas that may be airborne at
any time during dispersion of the cloud or plume.
Required Input Data
The model requires the following input parameter values'
• Molecular weight of the substance
• Normal boiling point of the discharged material (°F)
• Temperature of the substance in its container (°F)
• Ambient environmental temperature (°F)
• Vapor pressure of the substance after spill (mm Hg)
• Lower flammable limit of the gas or vapor (volume %)
• Atmospheric stability class (A to F)
• Wind velocity (mph)
• Vapor/gas emission rate (Ibs/minute)
• Duration of vapor/gas emission (minutes)
The program provides assistance (upon request of the user) in computing the molecular
weight of the discharged substance from its chemical formula via a procedure similar to that
descnbed in Section 2 8 of the guide
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats) Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
expenenced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can often (not always) be 20°F or more higher than the
ambient air temperature on a sunny day.
The user may request assistance in determining the vapor pressure of the discharged
substance at the point m time that the program asks for this parameter value See Section
12.28 below for a description of the Vapor Pressure Input Assistance Subprogram. The
model uses the vapor pressure in this case simply to confirm that flammable vapors can be
generated at specified temperatures in the external environment Thus, if the vapor pressure
of the gas or vapor is above one atmosphere (equivalent to 760 mm Hg), the value displayed
will be limited to a maximum vapor pressure of 760 mm Hg
12-69
-------
The lower flammable limit (LFL) of a gas or vapor was defined in Chapter 4 of this
guide. Values are always listed on material safety data sheets (MSDS) when appropriate and
available for the material of concern.
Assistance in selection of the appropriate atmospheric stability class and a wind
velocity consistent with this class is provided (upon request of the user) by reproduction of
the class selection chart (reproduced here in Table 12.8) The longest downwind hazard
distances along the centerlme of the wind direction are usually obtained when stability class
F is specified, and this is usually a good choice for general emergency planning purposes ~
as is a wind velocity specification of 4 5 mph However, be advised that there may be
exceptions to this general rule when an evaporating liquid pool is the source of hazardous
vapor emissions. Higher wind velocities typically produce greater evaporation rates, yet are
usually associated with atmospheric stability classes other than F Thus, the actual worst
case for toxic vapor dispersion hazard assessments may involve a different combination than
suggested above for some materials In the absence of more precise historical meterological
data for the region of concern, an atmospheric stability class of D together with a wind
velocity of about 10 mph can be assumed to evaluate threats under more typical atmospheric
conditions
The vapor or gas emission rate to the atmosphere as well as the duration of the
emission is usually computed by one or more of the preceding models on the Hazard
Assessment Model Selection Menu Any attempt to use the toxic vapor dispersion model
before these parameters have been computed will result in a warning to the user and a
question as to whether he or she wishes to continue the analysis
Model Results and Usage
Output of the model includes:
• Downwind hazard distance (feet)
• Maximum downwind hazard zone width (feet)
• Maximum weight of airborne gas (Ibs)
• Initial relative vapor/air density
• Type of model used for analysis
Two sets of results are provided for the first four items in the above list The first set is
based on a concentration that is 50% of the specified lower flammable limit (LFL) for the
gas or vapor in air. The second is based on the full value of the LFL
12-70
-------
The concentration of a gas or vapor at any specific point downwind will fluctuate about
an average value due to atmospheric turbulence even if all other factors that can influence
vapor dispersion phenomena remain unchanged Vapor dispersion models, including the
models in ARCHIE, are usually formulated to provide this average value as a function of
downwind location, which is perfectly acceptable for consideration of toxic gas or vapor
dispersion hazards. In the case of flammable gases or vapors, however, it is necessary to
make a distinction between that portion of a cloud or plume that can burn and that portion
that may explode, and this requires consideration of peak to average concentrations at
downwind locations. Without getting into more technical details, suffice it to say that a
cloud or plume has the potential to burn out to the boundaries of the area encompassed by a
gas or vapor concentration that is approximately one-half the LFL. The area subject to
explosion, however, is better estimated via use of the actual LFL value.
This procedure in ARCHIE actually consists of two separate and distinct models. If the
program decides that the gases or vapors released to the atmosphere are best treated as being
neutrally buoyant, it uses elements of the toxic vapor dispersion model to provide desired
results. Conversely, if the vapors are deemed to be best treated as being negatively buoyant
in air, it uses a specially formulated and simplified heavy gas model
Results of the model primarily provide emergency planning personnel with an
indication of the size of the hazard zone that may be subject to a vapor or gas cloud or plume
deflagration, and thus, indications of the number of people that may be killed or severely
injured, and the number and characteristics of buildings and other resources that may be
exposed to the flame The maximum weight of potentially explosive airborne gas or vapor
estimated by the model is typically utilized as an input parameter to the unconfined vapor
cloud or plume explosion model (Hazard Model Menu Option I).
Major Assumptions of the Methodology
Assumptions associated with use of the vapor dispersion model for neutrally buoyant
vapors or gases are discussed in Section 12 19 above Technical details and assumptions of
the heavy gas model are provided in Appendix B to this guide.
The decision as to which of the two models is to be used is based on a computation of
the ratio of the vapor-air mixture density to the density of pure air The computation
procedure leading to this relative vapor density is presented in Section 2 6 of the guide. The
concentration of gas or vapor in air is taken to be 100% by volume in the event of pressurized
gas releases
12-71
-------
The program assumes that the vapor or gas is best treated as being neutrally buoyant if
its initial relative vapor density has a value of 1 5 or below. Values above 1.5 lead to use of
the heavy gas model for negatively buoyant gases or vapors. Note that this version of
ARCHIE is incapable of taking the presence of airborne liquid aerosols into account and may
therefore occasionally err during this process.
Be advised and stay aware that there is at least one special case in which the type of
dispersion model(s) being employed by ARCHIE may underestimate downwind hazard zone
lengths. This involves scenarios in which a compressed liquefied gas under very high
pressure in its container vents a high velocity jet of gas and liquid aerosols at a high rate m
the downwind direction When the jet is forceful and the gas and aerosol mixture is
heavier-than-air, actual downwind travel distances to given concentrations in air may at times
exceed those predicted by ARCHIE.
1224 HAZARD MODEL MENU OPTION I: UNCONFINED VAPOR CLOUD
EXPLOSION MODEL
Purpose of Model
Although some flammable and therefore potentially explosive gases and vapors have a
greater propensity to explode than others upon ignition when at or above their respective
lower flammable limit (LFL) concentrations in the open environment, the threat of such an
explosion exists with a large number of hazardous materials. This threat increases for most
materials as the degree of confinement is increased.
The purpose of this particular model in ARCHIE is to evaluate the impacts of an
explosion involving an unconfined (or even partially confined) gas or vapor cloud or plume
in air.
Required Input Data
Input data and information required by the model include.
• Lower heat of combustion of the gas or vapor (Btu/lb)
• Yield factor for the explosion
• Weight of airborne flammable gas or vapor (Ibs)
• Location of the explosion relative to the ground surface
The heat of combustion of a material is the amount of heat generated when a unit
weight of the substance is burned under specified conditions The lower heat of combustion
ideally desired for model use is the amount of heat liberated when the material is burned m
oxygen at a temperature of 25°C (77°F) and the products of combustion, including any water
12-72
-------
that may form, remain as gases The input screen for the parameter provides guidance for
selecting a value when more precise data are unavailable. For informational purposes, it is
noted that the CHRIS Hazardous Chemical Data manual available at a cost of $41 in late
1988 from branches of the Government Printing Office contains the desired value for over
1200 hazardous materials This manual is widely used by fire service and other emergency
response personnel across the country and is very likely to be locally available No
hazardous materials library would be complete without the document
When a volume of gas or vapor in air burns and explodes, only a small fraction of the
energy in the cloud or plume actually contributes to the formation of shock or blast waves
that can damage people and property. This fraction is referred to as the yield factor and
vanes for different materials Upon request of the user, the program will provide guidance for
selection of an appropriate yield factor for the substance of concern when it is completely
unconfmed in the environment This guidance is reproduced in Table 129. Pay special
attention to the fact that the guidance only applies to situations in which the gas or vapor
cloud or plume is completely unconfined Yield factors can increase substantially when there
is a significant degree of confinement
The weight of flammable gas or vapor in air is usually computed by the vapor cloud or
plume fire model described above (Hazard Model Menu Option H) Any attempt to use the
unconfined vapor cloud explosion model before this weight has been computed will result m
a warning to the user and a question as to whether he or she wishes to continue the analysis
In addition, the user will be warned if the weight provided is less than 1000 pounds, since the
probability of a completely unconfined vapor cloud explosion (based on historical data) is
very low in such cases, except for a very few reactive substances and for situations in which
there is some degree of confinement pnor to ignition
As discussed in Chapter 5 of this guide, an explosion centered near the ground can
behave quite differently than an explosion at some point well above the ground surface (a so
called free air blast). Thus, the program asks the user whether a ground or elevated location
should be assumed Be advised that the damages caused by a groundlevel explosion will be
greater than those caused by a free air blast when all other factors are equal.
Model Results and Usage
The model produces a table which lists distances from the explosion center associated
with various degrees of injury and damage to people and property. It is important to realize
in the case of unconfined vapor cloud explosions that the center of the explosion could be
anywhere within the area subjected to gas or vapor concentrations at or above the LFLfor
the material of concern This area is usually defined by use of the vapor cloud or plume fire
model described in Section 12 23 (Hazard Model Menu Option H)
12-73
-------
TABLE 12.9
ASSISTANCE DISPLAY FOR EXPLOSION YIELD FACTOR
YIELD FACTOR SELECTION ASSISTANCE FOR COMPLETELY UNCONFINED
CLOUDS
EXAMPLES OF SUBSTANCES WITH YF = 0.03
Acetaldehyde
Acetone
Acrylonitrile
Amyl acetate
Amyl alcohol
Benzene
Butadiene
Butane
Butene
Butyl acetate
Carbon monoxide
Cyanogen
Cymene
Decane
Dichlorobenzene
Dichloroethane
Dimethyl ether
Dimenthyl sulfide
Ethane
Ethanol
Ethyl acetate
Ethyl amine
Ethyl benzene
Ethyl chloride
Ethyl formate
Ethyl propnonate
Furfural alcohol
Heptane
Hexane
Hydrocyanic acid
Hydrogen
Hydrogen sulfide
Isobutyl alcohol
Isobutylene
Isopropyl alcohol
Methane
Methanol
Methyl acetate
Methyl amine
Methyl butyl ketone
Methyl chloride
Methyl ethyl ketone
Methyl formate
Methyl mercaptan
Methyl propyl ketone
Naphthalene
iso - Octane
Pentane
Petroleum ether
Phthalic anhydride
Propane
Propanol
Propnonaldehyde
Propyl acetate
Propylene
Propylene dichlonde
Styrene
Tetrafluoroethylene
Toluene
Vinyl acetate
Vinyl chloride
Vmyhdene chlonde
Water gas
Xylene
EXAMPLES OF SUBSTANCES WITH YF = 0.06
Acrolein
Carbon disulfide
Cyclohexane
Diethyl ether
Divinyl ether
Ethylene
Ethyl nitrite
Methyl vinyl ether
Propylene oxide
EXAMPLES OF SUBSTANCES WITH YF = 0.19
Acetylene
Ethylene oxide
Ethyl nitrate
Hydrazine
Isopropyl nitrate
Methyl acetylene
Nitromethane
Vinyl acetylene
12-74
-------
Each of the hazard zone distances estimated by the program is associated with a
specific peak overpressure in the shock or blast wave and these in turn are related to specific
levels of expected damage using the information tabulated in Table 5 1 of this guide Table
12 10 illustrates the format of the general results obtained from ARCHIE but shows the peak
overpressures associated with each level of damage or injury instead of a specific distance
Results of the model provide emergency planning personnel with an indication of the
radii of the circular zones around the center of the explosion that may be subject to
explosion impacts of various levels of seventy, thus additionally providing indications of the
number of people that may be killed or severly injured, and the number, characteristics, and
level of damage that may be sustained by buildings, homes, and other resources that may be
exposed to the shock or blast wave
Major Assumptions of the Methodology
Be advised that model results assume the surrounding area is essentially flat and
without obstacles In actuality, potential reflections of the blast or shock wave from building
walls or the sides of other obstacles and surfaces may cause actual damage patterns to be
somewhat more erratic than those predicted by this generalized hazard assessment methodol-
ogy
12.25 HAZARD MODEL MENU OPTION J: TANK OVERPRESSURIZATION
EXPLOSION MODEL
Purpose of Model
The tank overpressunzation explosion model is used to evaluate the hazards resulting
from cases in which a sealed or inadequately vented tank or container may be internally
overpressunzed by a gas or vapor to the point of violent rupture This type of explosion is
descnbed and discussed more fully in Chapter 5
Required Input Data
The model requires knowledge of
• The shape of the tank or container
• Pressure at which the tank or container will rupture (psia)
• Gas or vapor volume in the tank (ft3)
• Ratio of specific heats (Cj/Cv) for the gas or vapor
• Ambient environmental temperature (°F)
• Temperature of the gas or vapor in the tank (°F)
12-75
-------
TABLE 12.10
EXAMPLE OUTPUT FROM EXPLOSION MODELS
Peak Overpressure
(psia)
0.03
0.30
1.00 - 0.50
1.00
8.00 - 1.00
2.00
3.00 - 2.00
12.2-2.40
2.50
4.00 - 3.00
5.00
7.00 - 5.00
10.0
29.0 - 14.5
EXPECTED DAMAGE
Occasional breakage of large windows under stress.
Some damage to home ceilings; 10% window breakage.
Windows usually shattered; some frame damage.
Partial demolition of homes, made uninhabitable.
Range serious/slight injuries from flying glass/objects.
Partial collapse of home walls/roofs.
Non-reinforced concrete/cinder block walls shattered.
Range 90-1% eardrum rupture among exposed population.
50% destruction of home brickwork.
Frameless steel panel buildings ruined.
Wooden utility poles snapped.
Nearly complete destruction of houses.
Probable total building destruction.
Range for 99-1% fatalities among exposed populations due to direct
blast effects.
Note: Output from the computer program shows Distance from Explosion in feet in place of
Peak Overpressures shown here for reference purposes. See Table 5.1 of guide for
additional information
12-76
-------
The first question asked by the program in cases where the answer has not been
previously specified is the shape of the tank or other container Valid choices are rectangular
tank, horizontal cylindrical tank, vertical cylindrical tank, and spherical tank This particular
model cannot be used for pipelines
Knowledge of the pressure at which the tank will rupture when overpressunzed will
require some detective work since tank construction standards vary widely for different types
of hazardous materials. The best bet is to query the owner or operator of the tank, and if
necessary, the company that initially constructed the vessel for this information
Prior to asking the user to provide the volume of the tank or other container that
contains compressed gas or vapor, an input parameter value that would otherwise require
several manual computations, the program asks if the user desires assistance in characterizing
the volume of the container and the weights, volumes, and physical states of its contents
Further details on the Tank and Container Contents Characterization Subprogram and its
general data requirements are provided in Section 1229 of this chapter. Use of the
subprogram may result in requests for a few informational items not listed above
The ratio of specific heats for the compressed gas or vapor is a thermodynamic
property related to the amount of heat necessary to increase the temperature of a unit weight
of gas or vapor by one degree under specified conditions Prior to request of this value, the
program will provide the user an opportunity to request assistance in selection of an
appropriate specific heat ratio using several generalized rules-of-thumb (reproduced in Table
12 7) that require user knowledge of the chemical formula of the hazardous material being
evaluated. Be advised that the results of this particular model are sensitive to the value
provided by the user. Accuracy of model results can and will be improved by provision of a
more precise value for the substance at or near the temperature in its container, but this is by
no means mandatory.
As discussed in Chapter 2 of this guide, many properties of hazardous materials are a
function of temperature, with one of the most important being the vapor pressure of the
substance (since this property will ultimately have a major effect on the magnitude of toxic
or flammable vapor dispersion threats). Consequently, for planning purposes, it is desireable
to select both tank and ambient environmental temperatures among the highest that may be
experienced during a typical year. In selecting these temperatures, note that the temperature
of the contents inside a metal tank can easily be 20°F or more higher than the ambient air
temperature on a sunny day
Model Results and Usage
12-77
-------
The tank overpressurization explosion model produces a table which lists distances
from the explosion center associated with various degrees of injury and damage to people
and property. Each of the hazard zone distances estimated by the model is associated with a
specific peak overpressure in the shock or blast wave and these in turn are related to specific
levels of expected damage using the information tabulated in Table 5 1 of this guide Table
12.10 illustrates the format of the general results obtained from ARCHIE, but shows the peak
overpressures associated with each level of damage or injury instead of a specific distance
Note that the model does not fully address the potential dispersion of high velocity tank or
container fragments that may result from the explosion and pose a threat to other containers
or stores of hazardous materials or other vulnerable resources
Results of the model provide emergency planning personnel with an indication of the
radii of the circular zones around the center of the explosion that may be subject to
explosion impacts of various levels of seventy, thus additionally providing indications of the
number of people that may be killed or severely injured, and the number, characteristics, and
level of damage that may be sustained by buildings, homes, and other resources that may be
exposed to the shock or blast wave.
Major Assumptions of the Methodology
Be advised that model results assume the surrounding area is essentially flat and
without obstacles In actuality, potential reflections of the blast or shock wave from building
walls or the sides of other obstacles and surfaces may cause actual damage patterns to be
somewhat more erratic than those predicted by this generalized hazard assessment methodol-
ogy.
Since most tanks or containers subject to this type of violent explosion or rupture are
likely to be on or near the ground, the model assumes this location for the explosion in all
cases. See Chapter 5 for a discussion of the difference between groundlevel and free air
explosions. The latter type of event will typically produce specified damages over lesser
distances than those predicted for groundlevel explosions
1226 HAZARD MODEL MENU OPTION K: CONDENSED-PHASE EXPLOSION
MODEL
Purpose of Model
This model is used to evaluate the hazards of solid or liquid explosive materials such as
nitroglycerine, TNT, dynamite, and a wide variety of lesser known substances with true
explosive properties. This type of detonation or explosion is more fully described and
discussed in Chapter 5.
12-78
-------
Substances most likely capable of detonation will be classified as explosives by the
DOT Explosives for which the model is applicable may also be formed at times during the
reaction of two or more substances which are not individually considered to have explosive
characteristics (see Chapter 7).
Be advised that substances classified as blasting agents by the DOT, despite then-
name, are considered to have a much lower likelihood of exploding, even in accidents
involving fire or impact See Chapter 8 for formal DOT definitions of blasting agents and
designated explosive materials
Required Input Data
The model requires only two input values, these being.
• Heat of combustion of the explosive material (Btu/lb)
• Weight of material at risk of exploding (Ibs)
The heat of combustion of a material is the amount of heat generated when a unit
weight of the substance is burned under specified conditions The lower heat of combustion
ideally desired for model use is the amount of heat liberated when the material is burned m
oxygen at a temperature of 25°C (77°F) and the products of combustion, including any water
that may form, remain as gases The input screen for the parameter provides guidance for
selecting a value when more precise data are unavailable. For informational purposes, it is
noted that the CHRIS Hazardous Chemical Data manual available at a cost of $41 from
branches of the Government Printing Office contains the desired value for over 1200
hazardous materials. This manual is widely used by fire service and other emergency
response personnel across the country and is very likely to be locally available No
hazardous materials library would be complete without the document
Model Results and Usage
The model produces a table which lists distances from the explosion center associated
with various degrees of injury and damage to people and property Each of the hazard zone
distances estimated by the program is associated with a specific peak overpressure in the
shock or blast wave and these in turn are related to specific levels of expected damage using
the information tabulated in Table 5 1 of this guide Table 12 10 illustrates the format of the
general results obtained from ARCHIE but shows the peak overpressures associated with
each level of damage or injury instead of a specific distance.
Results of the model provide emergency planning personnel with an indication of the
radii of the circular zones around the potential explosion site that may be subject to
explosion impacts of various levels of seventy, thus additionally providing indications of the
12-79
-------
number of people that may be killed or severly injured, and the numbei, characteristics, and
level of damage that may be sustained by buildings, homes, and other resources that may be
exposed to the shock wave
Major Assumptions of the Methodology
The model is the traditional TNT-equivalent procedure widely used for many years for
evaluation of high explosive detonation hazards A primary assumption, which is usually
quite valid for the types of explosives being considered, is that the blast or shock wave
produced by the explosion will dissipate in the same fashion as the explosion that would
occur if a weight of TNT having the same total energy of combustion were detonated.
Since the explosive material is most likely to be on or near the ground surface when the
explosion occurs, the procedure assumes this location for the explosion in all cases See
Chapter 5 for a discussion of the difference between groundlevel and free air explosions
The latter type of event will typically pioduce specified damages over lesser distances than
those predicted for groundlevel explosions
As in the case of the other types of explosions, be advised that model results assume
the surrounding area is essentially flat and without obstacles In actuality, potential
reflections of the blast or shock wave from building walls or the sides of other obstacles and
surfaces may cause actual damage patterns to be somewhat more erratic than those predicted
by this generalized hazard assessment methodology.
12.27 REMAINING OPTIONS ON THE HAZARD ASSESSMENT MODEL SELEC-
TION MENU
Option L: Review of Model Descriptions
Selection of this option permits the user to view several screens of text containing bnef
descriptions of the models available from the Hazard Assessment Model Selection Menu
Option M: Review of Model Selection Charts
This option permits the user to view the model selection charts that were previously
presented in Figure 12 2
Option N: Return to Main Menu
As the title suggests, selection of this option returns the user to the Mam Task Selection
Menu Be advised that a return to this menu will "close" the Accident Scenario File (ASF)
12-80
-------
being worked on while the Hazard Assessment Model Selection Menu is in use In order to
make this ASF file active again, it will be necessary to select Option B from the Main Task
Selection Menu and recall the file into active service
12 28 USE OF THE VAPOR PRESSURE INPUT ASSISTANCE SUBPROGRAM
Knowledge of the vapor pressure of a hazardous substance as a function of temperature
is an important prerequisite to an adequate evaluation of accident scenario results in most
cases There are likely to be many situations, even when some vapor pressure data points are
available for specific temperatures, however, when these data will be for temperatures other
that those needed by the program. Thus, the vapor pressure input assistance subprogram is
designed to facilitate characterization of the vapor pressure versus temperature relationship
for a hazardous substance regardless of whether detailed data are available or whether only
the data on a material safety data sheet are at hand Upon a request of assistance from the
user at one of the points in the program that vapor pressures are requested for specific
temperatures, the program will respond with a short menu offering three options, these
being
1
The user provides the vapor pressures at these temperatures
2. The user can provide whatever vapor pressure and other data are available
and permit the program to estimate the vapor pressures at the desired
temperatures
3 If available, the user can provide the coefficients for the Antome equation
for the substance. Further information is given when this option is selected
The first option is simply an escape mechanism for the user who arrives at this menu
unintentionally, and also, for the user who chooses one of the two following options, is not
satisfied with the results of the vapor pressure input assistance procedures, and desires to exit
the assistance subprogram
The second option takes advantage of any and all available vapor pressure data sets
(where a set is defined to be a specific vapor pressure and related temperature) for the
substance of interest Selection of this option results in a series of questions to the user
alternated with information screens The general order of events is
1. The user is asked if he or she has the vapor pressure for the material at a
temperature of 68°F (20°C), this being a vapor pressure available on
material safety data sheets for most volatile substances If the user answers
yes, he or she is next asked to choose the units in which the vapor pressure
will be provided Choices include pounds per square inch - absolute (psia),
12-81
-------
atmospheres (atm) , millimeters of mercury (mm Hg), Pascals (Pa),
Newtons per square meter (N/m2), and Bars (Note Definitions of these
units appear in Chapter 2 of this guide ) Once the value has been input in
the specified set of units, the program lists the vapor pressure provided in
all sets of available units and asks the user to confirm that the value was
entered properly.
2. The vapor pressure of any substance, as noted in Chapter 2, is always 760
mm Hg or the equivalent at its normal boiling point temperature at one
standard atmosphere The program will report this fact to the user simply
for informational purposes.
3. The subprogram will now have a least one data set and possibly two
Although the subprogram can generate a curve of vapor pressure versus
temperature from two points, the curve would not be very accurate Thus,
the subprogram continues with a question asking if the user has a vapor
pressure at another temperature A yes answer from the user will result in a
choice of units for the vapor pressure, a request for input of the vapor
pressure value, a choice of units for the related temperature, a request for
input of this temperature, and a screen seeking confirmation that the data
has been entered properly If three data sets are now available, the
subprogram will skip to item 5 below Otherwise, it will repeat this step
until it receives a no answer to its initial query.
4 If only one or two data sets are available at this point, the subprogram will
ask the user if the substance is flammable or combustible A yes answer
will result in requests for the lower flammable limit (LFL) of the substance
in volume percent and its flash point (closed cup preferred) in degrees
Fahrenheit if they are available. These data will be used to generate a data
set based on the assumption that the flash point of a substance is
theoretically equivalent to the temperature at which its vapor pressure
provides a vapor concentration equivalent to its LFL
5. If only one data set is available after all of the above steps, the subprogram
will inform the user that available data are insufficient and return him or her
to the opening menu of the subprogram If two data sets are available, the
user will be warned of the potential inaccuracy of vapor pressure estimates
and will be shown vapor pressure predictions for the normal boiling point
temperature of the hazardous material, the temperature of the material in its
tank or other container, and the ambient environmental temperature A
final question will ask if the predictions appear to be of acceptable
12-82
-------
accuracy. If three data sets are available, the subprogram will directly
proceed to a display of the above predictions and ask if results are
acceptable If they are, it will then display the Antoine equation coeffi-
cients derived from the input data and the temperature bounds in which the
equation should produce fairly accurate results.
Several excellent reference sources of chemical property data contain tables of vapor
pressure versus temperature for literally hundreds of substances, as well as a wide variety of
other data useful for hazard assessment purposes Among the more readily available of these
references m libraries and often in the possession of local chemical engineers and chemists
(including possibly the local high school chemistry teacher) are
Perry's Chemical Engineers' Handbook, McGraw-Hill Book Company,
New York, NY (various editions have different editors and publication
dates).
Weast, R C, et al, ed 's, CRC Handbook of Chemistry and Physics, CRC
Press, Boca Raton, FL (various editions have different publication dates)
The third option on the opening menu to the subprogram, as noted above, asks if the
user has access to the Antoine equation coefficients for the hazardous material of interest
For those who may not be familiar with this equation, be advised that it is a commonly used
equation with three coefficients When provided a temperature within the range of
temperatures for which it is applicable, it predicts the vapor pressure associated with this
temperature Each hazardous material has its own relatively unique set of coefficients that
can be found in various data sources. Two excellent sources of Antoine coefficients and
considerable other data for a large number of chemicals include.
Reid, R C, et al, The Properties of Gases & Liquids, McGraw-Hill Book
Company, New York, NY
Dean, J A, ed., Lange's Handbook of Chemistry, McGraw-Hill Book
Company, New York, NY.
Various editions of both of these books have differents publication dates, but all contain
valuable data and information The first book lists Antoine coefficients for almost 600
materials in its 4th edition. The second has coefficients listed for over 900 substances in its
most recent editions
The general form of the Antoine equation is
log(P) or In(P) = A - (B / (T + C))
12-83
-------
where A, B, and C are the three coefficients, T is the temperature at which the vapor pressure
is desired, and P is the vapor pressure (shown on the left side of the equal sign as the log or
natural log of the pressure).
Different reference sources use different units for the various coefficients and other
parameters associated with the equation, so the subprogram begins with a series of question
designed to determine the specific units for which the coefficients on hand are valid These
initial questions request*
• An indication of whether the left side of the equation provides the log or
natural log of the vapor pressure (P)
• An indication of whether the equation uses temperatures (T) in units of
degrees Celsius or Kelvin
• An indication of whether the vapor pressure predicted is in units of
millimeters of mercury (mm Hg), Pascals (Pa), Newtons per square meter
(N/m?), or Bars
• An optional specification of the temperature range over which the equation
is valid
Once the program has confirmed that the user has entered this information as he or she
intended, it proceeds to request the three equation coefficients It then predicts vapor
pressures for the normal boiling point temperature of the hazardous material, the temperature
of the material in its tank or other container, and the ambient environmental temperature A
final question asks if the predictions appear to be of acceptable accuracy.
12.29 USE OF THE TANK AND CONTAINER CONTENTS CHARACTERIZATION
SUBPROGRAM
A user of ARCHIE will very often have knowledge of the dimensions of a tank or other
container of the hazardous material of concern, the temperature and pressure of its contents,
and some idea of how full it might be on average or when m peak use The tank and
container contents characterization subprogram helps the user in translating this information
into the weights, volumes, and other characteristics of the container and its contents that are
required by ARCHIE, thus reducing or eliminating the need for a variety of manual
calculations. The subprogram is available from any input parameter screen in ARCHIE that
requests the weight of the hazardous material in the tank or the volume of the tank that
contains a compressed gas. Be advised that use of the subprogram is highly recommended
since denial of this offer of assistance creates a situation in which ARCHIE must accept input
data from the user without being able to ensure that all data are consistent
12-84
-------
Use of the subprogram is very straightforward First, it asks for several but not
necessarily all of the following input parameter values if these have not previously been
provided, including the:
• Shape of the tank or other container
• Length, width, height, and/or diameter of the tank as needed
Molecular weight of the contents
• Normal boiling point temperature of the contents (°F)
• Vapor pressure of the material in the container (psia)
• Temperature of the material in the container (°F)
• Ambient environmental temperature (°F)
• An indication of whether or not the tank contains any liquid
• Specific gravity of the liquid, if applicable
Once the above list of information has been obtained, the subprogram follows with
what might be the most important question of all (presuming the user has earlier indicated
there is liquid in the tank), this being "Which of the following do you know about the tank
contents?".
1 Amount of liquid in the container in gallons
2 Amount of liquid in the container in pounds
3 Amount of liquid in the container in cubic feet
4. Height of liquid in the container as measured from its bottom
5 Percentage of container filled with liquid
6 None of the above
Selection of one of these options and input of the appropriate answer leads to a screen
in which the dimensions of the tank and the characteristics of its contents are listed A final
question asks if the user wishes to repeat the procedure (in case some item or another looks
in error). A yes answer leads to a screen summarizing user input values and asking which
ones should be changed A no answer takes the user back to where he or she was in
ARCHIE prior to entering this subprogram
12.30 OTHER COMPUTER PROGRAMS
There are a number of other computer programs that may be of interest to emergency
planning and response personnel The EPA document entitled Identifying Environmental
Computer Systems for Planning Purposes (Preparedness and Prevention Technical Assis-
tance Bulletin #5) provides a checklist of computer system needs related to hazard analysis
and information on other available systems applicable to local planning efforts. The
checklist addresses a variety of systems for air dispersion and other environmental modeling,
for access to chemical properties data, and for emergency response planning. Note that
12-85
-------
different systems, particularly those associated with consequence analysis, may provide
differring results due to varying assumptions made during model formulation by system
developers.
12-86
-------
13.0 FORMULATION OF A PLANNING BASIS
13.1 INTRODUCTION
Chapter 10 of this guide provides guidance pertaining to the identification and
characterization of those activities or operations within a specific community or jurisdiction
that involve the use, storage, or transportation of hazardous materials and which therefore
pose some degree of hazard or risk to the public. Chapter 11 of the guide provides the data,
methods, and procedures necessary to estimate the probability of an accident associated with
each of these activities or operations. In addition, the chapter provides an opportunity in
many cases to assess the probabilities associated with different levels of accident seventy and
discusses how to analyze scenarios with consequence analysis procedures. Chapter 12
continues with a discussion of the ARCHIE program and its individual models that may be
used to evaluate many of the specific consequences of a spill, discharge, fire, or explosion
involving hazardous materials
Planning personnel who apply the procedures presented in the referenced chapters will
obtain a great deal of knowledge and insight into the specific threats facing their jurisdiction
from hazardous materials. Some may wish to use this knowledge directly in an attempt to
plan for all possible scenarios, and are not discouraged from .doing so if the time and
resources available are sufficient for this endeavor. For all intents and purposes, the entire
set of accident scenarios identified and considered during the hazard analysis will m
aggregate become the basis for their planning effort and they may proceed to Chapter 14
Those who do not have the resources available to plan for every possible contingency,
or who have other threats competing for attention, may wish to limit then- planning efforts for
hazardous materials by focusing on the most important situations This can be accomplished
by screening accident scenarios to identify those that have a reasonable likelihood of
occurrence in the foreseeable future and/or which may have significant consequences in the
absence of an organized, rapid, and effective response effort It is the purpose of Chapter 13
to assist these latter personnel in such a screening effort via a simplified risk analysis process.
The set of scenarios that remain after this process will then provide a planning basis for their
jurisdiction
13.2 DEFINITION OF ANNUAL ACCIDENT PROBABILITY CATEGORIES
Each of the accident scenarios identified during prior efforts should first be classified
into five categories based on the annual probability determined for the scenario using the
methods described in Chapter 11 of this guide The categories of interest include-
-------
• Common accidents,
• Likely accidents,
• Reasonably likely accidents,
• Unlikely accidents, and
• Very unlikely accidents
Common accidents are defined heiein as events expected to occur one or more tunes
each year on average. Such incidents have occurred in the locality of concern in the past and
are likely to reoccur with some regularity.
Likely accidents are defined herein as events expected to occur at least once every 10
years on average according to available statistics. Given the rapid increase in production and
usage of hazardous materials in the last 10 to 20 years, this type of accident is not one that
has necessarily occurred in the past in any given community, yet there is a 10% or greater
chance that it can occur in any given future year based on current levels of activity.
Reasonably likely accidents are events predicted to occur between once every 10 years
and once every 100 years on average. There is somewhere between a 1% and 10% chance of
such an event occurring in the locality of concern in any given year for the identified
scenario. In that similar scenarios involving different materials may also occur, the overall
likelihood of an event requiring a certain type of emergency response could be much higher.
(Note: This is equally true for the other categones.)
Ujilikely accidents are events predicted to occur between once every 100 years and
once every 1000 years on average in a specific locale. They are unlikely to occur within the
foreseeable future or during the lifetime of current inhabitants. Chances of their occurring in
any given future year are less than 1% and possibly as low as 0.1% for each specific
scenario.
Very unlikely accidents are events predicted to occur less than once in 1000 years The
odds are at least 100 to 1 against such an event occurring in the next 10 years in a specific
locale. The chances of one occurring in any particular future year are less than one m a
thousand.
If one is only evaluating releases qualitatively, common and likely accidents may be
equated to "high", reasonably likely and unlikely accidents to "medium", and very unlikely
accidents to "low" probability categones.
13-2
-------
13.3 DEFINITION OF ACCIDENT SEVERITY CATEGORIES
It is also necessary to classify accident scenarios according to their seventy as defined
via application of consequence analysis procedures. To assist in this effort, accident
consequences are classified into the four categones defined below, these being*
Minor accidents,
Accidents of moderate seventy,
Major accidents, and
Catastrophic accidents
Minor accidents are specified herein as those with the potential to have one or more of
the following features.
• Low potential for senous human injuries.
• No potential for human fatalities.
• No need for a formal evacuation, although the public may be cleared from
the immediate area of the spill or discharge.
• Localized, non-severe contamination of the environment which does not
require costly cleanup and recovery efforts
• No need for resources beyond those normally and currently available to
local response forces
Accidents are specified herein as of moderate seventy when they have the potential to
have one or more of the following features:
• Up to 10 potential human fatalities.
• Up to 100 potential human injuries requinng medical treatment or observa-
tion.
• Evacuation of up to 2000 people
• Localized contamination of the environment requinng a formal but quickly
accomplished cleanup effort
• Possible assistance needed from county and state authorities.
13-3
-------
• Only limited need for specialized equipment, services, or materials for a
rapid and effective response.
Major accidents are specified herein as those with the potential to have one or more the
following features:
• Up to 100 potential human fatalities
• Up to several hundred potential human injuries requiring medical treatment
or observation.
• Evacuation of up to 20,000 people
• Significant contamination of the environment requiring a formal and
somewhat prolonged cleanup effort.
• Assistance needed from county, state, and possibly federal authorities
• Significant need for specialized equipment, services, or materials for a
rapid and effective response.
Catastrophic accidents are defined as those having the potential to have one or more of
the following features:
• More than 100 potential human fatalities
• More than 300 potential human injuries requiring formal medical treatment
• Evacuation of more than 20,000 people.
• Significant contamination of the environment requiring a formal, pro-
longed, and expensive cleanup effort to protect human health and the
environment.
• Assistance needed from county, state, and federal authorities
• Significant need for specialized equipment, services, or materials for a
rapid and effective response.
Which category a given scenario falls into can be determined by considering
consequence analysis results along with local maps and population data
13-4
-------
There are two points to be made about the accident seventy categones defined above.
The first is that they do not represent hard and fast rules What some people may perceive as
a minor or routine incident may seem to be a major event to others Do not hesitate,
therefore, to make any changes to the procedure if some point or another makes you
uncomfortable After all, the authorities in a major city are likely to deal with loss of life and
senous injuries due to various "non-natural" causes on almost a daily basis. It is understand-
able that they may have a different perspective on such events than the authorities of a small
and quiet town in a rural part of the country.
The second point to be made is that the accident seventy categones above are partially
defined in terms of potential deaths and injuries. It must be realized that consequence
analysis procedures essentially provide estimates of the area or zone that may be subjected to
harmful levels of airborne contaminants, thermal radiation, and/or explosion overpressure
Although people within such an area will have a definite potential to be killed or injured, a
large fraction may actually escape unharmed Reasons for this are varied and complex, but
think about how many times the evening news has shown a community devastated by a
tornado, with dozens of buildings and homes knocked flat, and only reported a death or two
(if any) and a handful of injuries
13.4 APPLICATION OF SCREENING GUIDELINES
Figure 13.1 presents a matrix of accident probability categones versus accident seventy
classes Each block of the matrix suggests a planning approach to be taken for accident
groups that meet cntena associated with the block. These approaches are only suggestions
and should be treated as such. Special circumstances may require special consideration on a
case by case basis. Specific guidelines should be worked out by each locale to best represent
the resources and relationships between organizations applicable to that community.
The matrix should be applied only at the level of government or industry that identified
and analyzed the planning basis scenanos for a particular locale or jurisdiction The reason
for this stems from the fact that levels of authonty above the local level typically have
responsibility for several individual junsdictions Since there is a higher probability that a
particular type of accident will take place within a large junsdiction than in a specific smaller
one within its borders (e g, a county vs. a town), a higher level of authonty may find it
necessary to plan for accidents that are too rare to worry about at a lower level This does not
mean, however, that all junsdictions should not work together in coordinating their response
to rare events Rather, it means that the higher level authority should probably take a
leadership role in such instances
13-5
-------
FIGURE 131
ACCIDENT FREQUENCY/SEVERITY SCREENING MATRIX
o
a
0)
CD
I-T-)
Severity
Reasonably
Likely
Comprehensive planning and preparedness are essentially
mandatory at the appropriate levels of government or industry
Comprehensive planning is optional and does not necessarily
warrant any major efforts or costs Give consideration to
sharing any necessary special response resources on a regional
basis
Comprehensive planning may be unwarranted and unnecessary
13-6
-------
The frequency (annual probability) has been reduced to three categories to reflect the
fact that the treatment of individual scenarios may produce low annual probabilities, yet the
chance of a similar accident involving one of a number of different hazardous materials may
be much higher. The original five categones may be used to prioritize accident scenarios
more precisely within the broader categones.
13.5 MOTIVATION FOR CONTINUED PLANNING
The EPA recently sponsored an intensive effort to determine the number and severity
of accidents involving hazardous material releases in the six-year period of 1980-1985. The
total number of deaths recorded was 138, for an average of 23.0 per year, for the entire
United States. The annual average expected for any jurisdiction of 100,000 people can
therefore be computed as being on the order of 0.01 per year. Given that the identification of
deaths due to all episodic hazardous material accidents is difficult due to current reporting
requirements, it is likely that the EPA underestimated the true extent of the problem.
Nevertheless, the fatality rate would still be relatively minor even if multiplied by 10 or even
100.
These figures can be compared to annual mortality statistics in the United States to put
the overall problem of hazardous materials into the proper perspective. To help do this,
Table 13.1 presents annual mortality rates for a variety of natural and accidental causes of
death using statistics for recent years. The actual number of deaths that may be expected
each year on average due to any specific cause in any particular jurisdiction can be obtained
by multiplying the annual mortality rate for that cause by the population of concern.
Given the above findings one might wonder why there is such widespread concern
about the dangers of hazardous materials. The answer to this question is actually quite
complex. Part of it involves the fact that recent major accidents in other countries have
demonstrated the potential for a single accident to cause hundreds if not thousands of deaths
and injuries. Although the safety record of the U S. chemical industry has generally been
excellent, in terms of loss of life, people here (including government agencies) have naturally
felt a need to reassess safety standards and ensure our country is prepared to deal with future
emergencies. Yet another part of the answer involves the realization that hazardous material
accidents differ greatly from more typical accidents and require special preparations for an
effective response Indeed, it is precisely a lack of knowledge about chemical threats facing
a community, and a lack of specific preparations to counter these threats, that can lead to a
disaster that may otherwise have been prevented
Finally, it must be appreciated that the general public has a greater fear of threats that
can cause multiple fatalities, threats they may not fully understand or have any control over,
and threats due to activities that do not provide a direct benefit on an individual basis. The
13-7
-------
TABLE 13.1
ANNUAL INDIVIDUAL MORTALITY RATES FOR
NATURAL AND ACCIDENTAL CAUSES OF DEATH
Cause of Fatality
All Diseases
Heart Disease
Cancer
Cercbrovascular Disease
Pneumonia
Diabetes
All Accidents
Motor Vehicles
Falls
Drowning
Fires, Burns
Natural Hazards and Environmental Factors
Cataclysm (tornado, flood, earthquake, etc.)
Excessive Heat
Excessive Cold
Lightning
Risk of Fatality
(per 100,000 people)
830
320
190
64
28.3
15.5
39
19
5.0
2.2
2.1
0.8
0.09
0.09
0.40
0.04
Source: Accident Facts. 1988 Edition, National Safety Council
13-8
-------
reasons why tens of thousands of highway deaths are tolerated each year partially stem from
the facts that they usually occur one or two at time, individuals have control over whether
they drive or nde in automobiles, and because automobiles are perceived to provide a
substantial benefit to each individual.
13-9
-------
-------
14.0 USE OF HAZARD ANALYSIS RESULTS
IN EMERGENCY PLANNING
14.1 INTRODUCTION
Planning personnel who have followed the preceding guidance of this document should
have gained considerable insight into the true problems and nsks associated with hazardous
materials in their respective jurisdictions Additionally, they should have an understanding
of the probability, nature, and likely consequences of potential accidents. It remains to
discuss how this knowledge can be best utilized during plan preparation to ensure prompt,
efficient, and effective response to actual emergencies
The discussions that follow are intended to supplement guidance provided by the
National Response Team in the Hazardous Materials Emergency Planning Guide (NRT-1)
That particular publication discusses how to organize the overall emergency planning process
and provides a suggested outline for the resulting plan, an outline that contains numerous
important topics not directly associated with hazard analysis results but vital to the
completion of a comprehensive emergency plan In contrast, this chapter mostly limits itself
to planning issues directly related to findings of a hazard analysis and has the objectives of 1)
providing more detailed guidance in this topical area than found in NRT-1; and 2) bridging
the gap between this guide to hazard analysis and other more comprehensive publications
devoted to emergency planning and/or emergency response to emergencies.
Readers are advised that this chapter focuses on planning for significant incidents
involving spillage or discharge of a hazardous material and not the minor types of spills or
leaks that are relatively common and of a routine nature. In addition, it should be noted that
chapter contents are only intended to give readers a strong push in the right direction.
Decisions will be necessary on a case by case basis as to whether any particular planning
activity is desired or considered warranted by local conditions Special or unique circum-
stances may require planning efforts and activities not addressed by this chapter.
Finally, be advised that the chapter is directed to planning personnel in state and local
governmental agencies. Although much of the guidance and discussion presented in the
chapter is also applicable to emergency planning for specific industrial sites, such sites are
likely to have additional planning needs not addressed herein For the benefit of industrial
planning personnel, a later section of the chapter references specific publications better suited
to the needs of industry for development of m-plant emergency response plans.
-------
143 ORGANIZATION OF THE CHAPTER
Chapter 14 addresses 43 separate topics associated with emergency response planning
in 12 subject areas, with Table 14.1 serving as an overall index. These topics are assigned
unique "item" numbers for reference and organization purposes
Each topic is presented and discussed in an individual section of the chapter using a
standardized format Each of these sections denotes the circumstances under which the
planning item is applicable, states the planning goal to be achieved, suggests one or more
action items for accomplishment of this goal, and ends with a brief discussion of the overall
topic. Planning personnel should refer to the Hazardous Material Emergency Planning
Guide for advice on how to organize results of their planning activities into their final written
plan.
14.3 ADDITIONAL SOURCES OF PLANNING GUIDANCE AND INFORMATION
Additional details on what to plan for, how to plan, and how to manage response to an
overall incident involving spillage or discharge of a hazardous material can be found in
numerous other publications.
Publications of the Federal Emergency Management Agency
Given that FEMA is the primary agency of the federal government having responsibili-
ty for planning for all types of emergencies with the potential to threaten public health and
safety in the United States, and that the agency employs numerous personnel who are experts
in this field, it is not surprising to find that FEMA has published several documents that can
be extremely useful to state and local planning agencies as well as industrial facilities.
Indeed, it is strongly recommended that planning personnel obtain, review, and make use of
at least the following three documents while planning for emergencies related to non-ra-
dioactive hazardous materials. Chapter 1 provides information on how these documents may
be obtained, though it is highly likely that most governmental emergency preparedness
organizations (particularly civil defense agencies) will already have copies of the first two
publications on hand.
• Guide for Development of State and Local Emergency Operations Plans
• Guide for the Review of State and Local Emergency Operations Plans
• Hazardous Materials Contingency Planning Course (student manuals)
14-2
-------
The planning course is designed to be presented by a trained instructor and involves the
viewing of several video tapes in the classroom. Nevertheless, several of the nine separate
student reference manuals contain truly excellent and highly detailed information and
guidance on important topics that could only be allocated a few paragraphs of general
discussion in this chapter.
Publications of the Disaster Research Center
Individuals wishing to obtain a more formal and indepth "education" with respect to
problems and issues associated with emergency planning and response are advised to request
a publication list from the Disaster Research Center, University of Delaware, Newark,
Delaware, 19716 (telephone 302-451-6618). A representative sample of some excellent
books, reports, and articles of potential value to both public and industrial planning personnel
includes.
• Tierney, K.J., A Primer for Preparedness for Acute Chemical Emergen-
cies, Book and Monograph #14,1980.
• Quarantelh, E L, Organizational Behavior in Disasters and Implications
for Disaster Planning, Report Senes #18,1985.
Quarantelh, E.L, and Gray, J., First Responders and Their Initial
Behavior in Hazardous Chemical Transportation Accidents, Preliminary
Paper #96,1985
• Gray, J, Three Case Studies of Organized Responses to Chemical
Disasters, Misc Report #29,1981.
• Quarantelli, EL, People's Reactions to Emergency Warnings, Article
#170,1983.
• Quarantelli, EX., et al, Evacuation Behavior and Problems: Findings
and Implications from the Research Literature, Book and Monograph
Senes #18,1984
• Quarantelh, E.L, et al, Evacuation Behavior: Case Study of the Taft,
Louisiana Chemical Tank Explosion Incident, Misc Report #34,1983
Publications Oriented Towards Industrial Emergency Planning
Guidance documents particularly suited to the task of emergency planning for the
protection of employees and property within the boundaries of individual industrial facilities
include:
14-3
-------
• Site Emergency Response Planning, Chemical Manufacturers Association,
1986. See Chapter 1 for information on obtaining publications of this
organization.
• Newton, JE., A Practical Guide to Emergency Response Planning,
Pudvan Publishing Co, 1935 Shermer Road, Northbrook, Illinois, 60062,
1987 (telephone 312-498-9840)
Other Books and Reports
Dozens of books and manuals pertaining to planning or management of hazardous
material emergencies have been written over the years, with literally a flood of such
publications (some good, some not so good) appearing in the aftermath of the Bhopal
tragedy. A sample of publications worthy of special mention (who due apologies to authors
whose works were not identified or reviewed) includes
• American Petroleum Institute, Developing a Highway Emergency Re-
sponse Plan for Incidents Involving Hazardous Materials, API Recom-
mended Practice 1112, available from the API, 1220 L Street Northwest,
Washington, D.C., 20005, November 1984
• Bennett, G R, et al, Hazardous Materials Spills Handbook, McGraw-Hill
Book Company, New York, 1982.
• Cashman, J.R, Hazardous Materials Emergencies Response and Control,
TECHNOMIC Publishing Company, 851 New Holland Avenue, Box 3535,
Lancaster, Pennsylvania 17604,1983
• International Association of Fire Chiefs, Fire Service Emergency Manage-
ment Handbook, available from IAFC Foundation, 101 E Holly Ave -
Unit 10B, Sterling, VA 22170, January 1985
Omohundro, J.T, Oil Spills: A Public Official's Handbook, Report No
PB80-184351, prepared for the National Oceanic and Atmospheric Admin-
istration, available from the National Technical Information Service,
Springfield, VA 22151, March 1980.
• Shaver, D K., and Berkowitz, R L, Guidelines Manual: Post Accident
Procedures for Chemicals and Propellants, Report No AFRPL
TR-82-077, prepared for the Air Force Rocket Propulsion Laboratory and
the U.S. Department of Transportation, available from the National
Technical Information Service, Springfield, VA 22151, January 1983
14-4
-------
Shaver, D K., Berkowitz, R.L., and Washburne, P.V., Accident Manage-
ment Orientation Guide, Report No. AFRPL TR-82-075, prepared for the
Air Force Rocket Propulsion Laboratory and the U.S. Department of
Transportation, available from the National Technical Information Service,
Springfield, VA 22151, October 1983.
Smith, A J., Jr, Managing Hazardous Substance Accidents, McGraw-Hill
Book Company, New York, 1981.
14-5
-------
TABLE 14.1
INDEX TO PLANNING ITEMS
Item No. Subject Page
Notification
Nl Initial Notification of Spills 14-8
N2 Notification of Response Organizations and Public Authorities 14-9
N3 Facilities Requiring Special Notification 14-11
N4 Notification of Water Users 14-12
N5 Notification of Water Treatment Plants 14-13
N6 Notification and Shutdown of Electric and Gas Utilities 14-14
N7 Notification and Control of Air, Rail, and Waterborne traffic 14-15
Command and Communications
CC1 Establishment and Staffing of Command Posts 14-16
CC2 Establishment of Emergency Communications Systems 14-19
CCS Formulation of Response Objectives and Strategy 14-21
CC4 Ensuring Health and Safety at Incident Sites 14-23
Evacuation
EV1 Designation of Decision Responsibility for Public Protective Actions 14-26
EV2 Criteria for Evacuation and Shelter-in-Place Decisions 14-27
EV3 Public Alert/Notification/Lnstruction Procedures 14-29
EV4 Transportation Functions 14-31
EV5 Care and Shelter of Evacuees 14-33
EV6 Security of Evacuated Hazard Zones 14-35
EV7 Movement and Care of Domestic Livestock and Pets 14-36
Fire Response
FF1 Special Firefighttng Equipment and Materials 14-38
FF2 Identification of Water Sources in Rural Areas 14-39
Health Care
HC1 Establishment of Special Rescue Squads 14-40
HC2 Provision of Ambulance Services 14-42
HC3 Establishment of a Mass Casualty Plan 14-43
Personal Protection
PP1 Availability of Respiratory Protective Devices 14-46
14-6
-------
TABLE 14.1 (Continued)
INDEX TO PLANNING ITEMS
Item No. Subject Page
Personal Protection
PP2 Availability of Special Protective Clothing 14-48
PP3 Decontamination of Exposed Protective Clothing and Other Response 14-50
Equipment
Public Relations
PR1 Public Relations in Emergency Situations 14-51
Spill Containment and Cleanup
SCI Plugging/Stopping of Leaks 14-54
SC2 Suppression of Hazardous Gas or Vapor Releases 14-56
SC3 Containment of Spills of Liquids or Solids on Land 14-59
SC4 Cleanup of Spills of Liquids or Solids on Land 14-61
SC5 Containment of Spills into Water Bodies 14-63
SC6 Cleanup of Spills into Water Bodies 14-65
SC7 Support Services for Field Response Forces 14-66
SC8 Maintenance of Apparatus and Equipment 14-67
Spill Documentation
SD1 Documentation of Response Activities and Costs 14-68
Spill Monitoring
SMI Momtonng of Atmospheric Conditions 14-69
SM2 Momtonng of Contaminant Concentrations 14-70
Post-Spill Recovery
SRI Provision of Alternate Water Supplies 14-71
SR2 Cleanup of Dead or Contaminated Livestock or Wildlife 14-72
SR3 Post-Incident Testing for Contamination 14-73
SR4 Structural Inspections after Fires or Explosions 14-74
SR5 Provision of Post-Incident Recovery Services 14-75
Training
TR1 Training of Response Personnel 14-76
Waste Disposal
WD1 Disposal of Hazardous Wastes 14-77
14-7
-------
ITEM#N1
Topic: Initial Notification of Spills
When/Where Applicable. Any locale with hazardous material spill or discharge poten-
tial
I
Planning Goal: Establishment of central contact pomt(s) for initial notification
of spills to facilitate response plan activation
Action Items:
• Identify the 24-hour telephone number(s) that is (are) to be called by
members of the public, industry personnel, and government employees to
report a spill or discharge or any situation that could result in such an event
in the near future.
• Identify and list the information that should be requested from the caller to
facilitate initial assessment of the gravity of the situation
• Ensure that all parties have been advised of reporting procedures
Discussion:
Ordinary citizens are likely to call local police or fire departments if they observe an
accident, and it is reasonable to assume that most jurisdictions will not discourage the
practice. This does not mean, however, that one of these organizations must always serve as
the central point of contact for coordinating communications during hazardous material
emergencies. A better choice in some jurisdictions may be a predesignated emergency
command post or emergency operations center at either the local or county level that has
regular telephone, "hot" line, and/or radio communications links with response forces, public
authorities, and major industrial complexes
Li localities with major fixed-site potential spill sources, it is vital that all facility
operators know exactly when and where to call to report an incident to local government
authorities, and that all public officials in the area are in agreement as to required notification
procedures. Bluntly stated, the planning group should make every effort to resolve "turf
battles" and interagency squabbles as well as the real (yet sometimes political) needs of
elected officials.
14-8
-------
ITEM#N2
Topic. Notification of Response Organizations and Public Authorities
When/Where Applicable Any locale with hazardous material spill or discharge poten-
tial
Planning Goal. Rapid activation of emergency response forces at a level
consistent with the known gravity of the situation
Action Items
• Provide the central contact point(s) with instructions and procedures
relating to alert or call-out of response forces, including the individual or
his/her alternates with command responsibility for spill response in the
jurisdiction of concern
Discussion:
The watch-stander or dispatcher who first receives notification of an accident involving
hazardous materials should have fairly detailed instructions with respect to who should be
notified of the event and/or who should be asked to respond Minor incidents may simply
require dispatch of a fire department to the scene. More significant events may necessitate
call-out of additional forces and notification of county, state, and federal authorities In any
case, it should be clear to all as to who will be alerted under varying circumstances and who
will be in charge of response actions during various phases of an emergency situation.
Communities that frequently experience hazardous material emergencies of a minor
nature but are only rarely faced with more significant events may wish to consider a staged
response For example, depending on the seventy of the situation as described by the initial
caller, various levels of response might be established, thus avoiding the immediate need to
call out forces in strength for all incidents Personnel amvmg at the scene could, of course,
request additional assistance and thereby raise the level or stage of response, much as
additional alarms might be sounded during major fires or the threat thereof
One strategy to be considered for establishing levels of response classifies responses
into three levels as follows*
Level 1 An incident which can be controlled by the first response agencies and
does not require evacuation of other than the involved structure or the immediate
outdoor area The incident is confined to a small area and does not pose an
immediate threat to life or property
Level 2 An incident involving a greater hazard or larger area which poses a
potential threat to life or property and which may require a limited evacuation of
the surrounding area
14-9
-------
Level 3: An incident involving a severe hazard or a large area which poses an
extreme threat to life and property and will probably require a large scale
evacuation; or an incident requiring the expertise or resources of county, state,
federal, or private agencies/organizations
If military explosives can be found in the region of concern at times, the plan should
identify the nearest military base that can provide assistance with these materials Similarly,
where radioactive materials may pose a threat, it should identify the nearest radiological
emergency assistance team established by the Department of Energy or the state (see NRT-1
for DOE regional contacts). Where etiologic materials (those posing biological or biomedi-
cal hazards) may be encountered, it is well to list the emergency information telephone
number for the Director of the Center of Disease Control in Atlanta, Georgia, this being
404-633-5313. Other important telephone numbers at the national level include'
National Response Center- 800-424-8802 (202-426-2675 or 202-267-2675
in Washington, DC area), for notification, information, and assistance
involving agencies of the federal government
• Chemical Transportation Emergency Center (CHEMTREC):
800-424-9300, for information and assistance from industry as coordinated
by the Chemical Manufacturers Association
14-10
-------
ITEM#N3
Topic- Facilities Requiring Special Notification
When/Where Applicable: Where a major incident may occur in a location that threatens
schools, day care centers, hospitals, nursing homes, universi-
ties, prisons, and similar facilities serving large groups of
people with needs for special transportation, protection, or
handling.
Planning Goal: To ensure that the above facilities receive the earliest possible
notification of the need to shelter-in-place or evacuate their
inhabitants
Action Items:
• Identify and obtain the telephone numbers and names of key supervisory
personnel in facilities that meet the above criteria.
• Ensure that the emergency response plan provides for direct notification of
these facilities on a prompt basis
• Ensure that the emergency response plan contains a procedure for notifica-
tion of these facilities during penods of interrupted telephone service
• Encourage facility operators to establish emergency evacuation plans
coordinated with those of the community.
• Ensure that the person(s) responsible for notification of these facilities
document then* actions during an emergency.
Discussion:
Time can be critical during some hazardous material emergencies The above action
items ensure that facilities requiring special attention are notified at the earliest possible
indication of a threat, thus permitting them an early start to evacuating or otherwise
protecting their inhabitants.
14-11
-------
ITEM#N4
Topic: Notification of Water Users
When/Where Applicable: Any locale that may experience a spill of a hazardous material
into a body of water.
Planning Goal: To ensure that industrial, agricultural, public water supply, and
other users of water resources are quickly notified of potential
water contamination in the aftermath of a spill
Action Items*
• Compile a list of those companies or facilities that extract water from water
bodies in the area of concern, together with a list of appropriate contacts
and telephone numbers at these facilities
• Establish a mechanism by which these facilities may be quickly informed
of potential water contamination in the event of an upstream hazardous
material spill.
Discussion*
Numerous facilities intake water from nearby water bodies for industrial or food
processing purposes, farm irrigation, drinking supplies, and so forth Entry of a toxic,
flammable, or corrosive material into their water intakes can contaminate food or drinking
water, damage equipment, ruin products, and possibly even cause a fire or explosion
These facilities may be identified during the survey process discussed in Chapter 10, by
a special survey of facilities along the "waterfront", or possibly, by a search of local, state, or
federal permit records (presuming some sort of permit was required for them to intake water
from the water bodies in question)
Note that it may not be necessary for public authorities to call a long list of facilities
themselves It is entirely feasible to establish a "waterway warning network" in which each
person receiving a call on behalf of a facility is given responsibility to call two or more other
facilities in the downstream direction This would mean that public officials would only
have to make one or two telephone calls to start the process
In undertaking this planning task, realize that spills with the potential to adversely
impact water quality, particularly spills into rivers, need not actually occur in the jurisdiction
of concern Indeed, they could occur at locations many miles upstream from junsdictional
boundaries.
14-12
-------
ITEM#N5
Topic. Notification of Water Treatment Plants
When/Where Applicable: Where spillage of a hazardous material may enter a sewage or
drainage system leading to a municipal treatment plant
Planning Goal. To ensure that the treatment plant is warned as soon as
possible of the event
Action Items
• Identify the 24-hour telephone numbers and names of key supervisory
personnel in the subject facilities.
• Plan to have these personnel notified immediately if hazardous materials
enter the waste streams to their plant
• Identify sewer shut-off points for the containment of hazardous materials
that may leak or flow into sanitary and storm sewers
Discussion:
The sudden appearance of a toxic or flammable substance can cause many problems
and senous hazards at a treatment plant Early notification may help prevent or mitigate
adverse impacts.
A study sponsored by the EPA may be of use to biological wastewater treatment plants
in assessing the effects of potential hazardous materials spills and developing emergency
plans. The reference is:
Bnnsko, G.A., et al, Hazardous Material Spills and Responses for Municipali-
ties, Report No PB 80214141, available from the National Technical Information
Service, Springfield, VA 22161, July 1980.
Note that this report is solely devoted to wastewater treatment plant problems. Its title is
somewhat misleading in that it suggests broader coverage of spills and related response
actions.
See Item #SC3 for additional information and discussion of problems associated with entry
of flammable liquids into storm drains
14-13
-------
TTEM#N6
Topic: Notification and Shutdown of Electric and Gas Utilities
When/Where Applicable: Wherever a large explosion or potential explosion may
necessitate shutdown of electric power or natural gas distribu-
tion systems in an area.
Planning Goal: To identify who and where to call to have the above utilities
shutdown on a rapid basis.
Action Items:
• Contact local utility companies and establish a mechanism to accomplish
the planning goal.
Discussion:
Major explosions may be caused by or may themselves cause leaks or ruptures of
natural gas distribution systems. Additionally, they can cause breaking of power lines and an
electrocution hazard to those who might make contact with any "downed" lines. In either
case, there may be circumstances in which it is desired to shutdown natural gas or electric
power systems rapidly in an area.
14-14
-------
ITEM#N7
Topic: Notification and Control of Air, Rail and Waterborne Traffic
When/Where Applicable: Any locale in which a hazardous material accident may
threaten airplanes, helicopters, trains, barges, ships, recreation-
al boaters, and other non-highway traffic not normally aware
of local conditions.
Planning Goal: To identify the means to warn aircraft, trains, and waterborne
traffic of a hazard and to keep them away from the hazard
zone.
Action Items*
• Identify air traffic control facilities, railroad dispatchers, and Coast Guard
or harbor master facilities that have the ability to warn and control the
movement of the subject traffic in the area or jurisdiction of concern
Discussion:
The planning goal and action items are mostly self-explanatory. It is only necessary to
add that a train accident on a heavily travelled segment of track may require immediate
dispatch of personnel to locations where they can attempt to flag down and stop other trains
approaching the accident site. Overflight of any incident site should be prohibited except
with permission from the Federal Aviation Administration (FAA) and on-scene authorities.
Where available, helicopters having public address systems can facilitate warning of traffic
and even the general public.
14-15
-------
ITEMtfCCl
Topic: Establishment and Staffing of Command Posts
When/Where Applicable: Wherever a coordinated, large-scale response may become
necessary due to a major spill or discharge of hazardous
materials.
Planning Goal: To quickly establish one or more command posts from which
the emergency response may be directed and coordinated.
Action Items:
• Select an appropriate location (and possible alternate) for rapid establish-
ment of a primary command post or emergency operations center.
• Plan for a field command post near the site of the emergency from which
spill containment and countermeasure operations may be conducted when
necessary.
• Designate the individuals who should immediately report to each site in the
event of a major emergency.
• Establish a "check-in" location where key officials can be "logged in" and
their movements tracked once they appear on the scene
• Establish an ID system to control and track entry and movement of public
authorities and emergency crews.
• Equip sites with the office equipment, maps, data sources, communication
systems, and other supplies and resources necessary for command and
control of response activities.
• Provide security and access control at vital sites such as the emergency
operations center, communications center, media center, emergency supply
center or depot, and the incident site itself.
Discussion:
A complex emergency requires coordination of numerous activities beyond spill
containment and countermeasure efforts There can therefore be benefits to establishment of
both central and field command posts. The former, while in close communication with the
field site, can handle relations with the press, evacuation operations, contacts with the public
and outside government agencies, procurement of necessary supplies and resources, and a
variety of other details The field site can then concentrate on its own key mission of
directing hazard containment and control operations.
14-16
-------
The central command post should ideally be assigned representatives of all agencies
and organizations with a role in emergency response, as well as support personnel including
secretaries, clerks, radio operators, messengers, and so forth. This will facilitate communica-
tions, decision-making, and conduct of necessary efforts.
Senous consideration should be given to establishing a "check-in" location at each
command post, for requiring all individuals to "sign in" and "sign out" (particularly at the
incident site), and to issuing ID cards and/or badges to personnel authorized to pass through
roadblocks and enter controlled areas. Color-coded badges or items of clothing may be used
to designate different levels of access and/or on-scene authority
Note that the establishment of a local command organization for management of
emergencies is a related prerequisite more fully discussed in the Hazardous Materials
Emergency Planning Guide and the Hazardous Materials Contingency Planning Course
manuals referenced earlier in this document Emergency plans should contain a clear and
concise summary of primary and support responsibilities for Command and Control,
Alerting and Notification, Communications, Public Information, Accident Assessment,
Public Health and Sanitation, Social Services, Fire and Rescue, Traffic Control, Emergency
Medical Services, Law Enforcement, Transportation, Public Protective Actions, Exposure
Control, and Public Works functions, among others. Block diagrams should illustrate the
relationships among the various response groups, each of which may require assignment of a
specific manager under the overall direction of an individual assigned overall responsibility
and authority for command of emergency operations The command and control team and
its written plan must provide a capability for 24 hour protracted operations (requiring the
establishment of work shifts), definition of decision-making processes in at least general
terms, availability of sources of assistance (including specialized experts) to aid deci-
sion-making, and establishment of a means to determine operational readiness. Possible
staging areas should be identified for gathering and dispatch of response forces as necessary.
One or means must be available to obtain detailed and accurate information on the hazards
and properties of any and all chemicals involved in an accident on a rapid basis
Keep in mind during planning for the above activities that areas outside established
hazard zones will continue to require some degree of police, fire, rescue, ambulance, health
care, and public works services.
Give consideration to identifying sources of video and telephoto equipment that can be
used to view particularly hazardous accident sites from a safe distance on a continuous basis
at command, control, and media centers. A closeup view of the site on a television screen
can be extremely helpful to all parties to the response action in evaluating and choosing the
appropriate course of action, as well as satisfy media requests for photo coverage and many
of the information needs of government officials on scene It also enables experts to provide
guidance to response personnel approaching the site to undertake fire control, leak plugging,
spill containment, and/or spill cleanup efforts.
14-17
-------
One of the first steps of a response action must be restriction of access to the spill site
and other hazardous areas. There is a not-so-old saying that states "if you want to draw a
very large crowd instantly, in the middle of nowhere, dump something lethal on the ground
and set it on fire." Experienced emergency coordinators will attest to the validity of this
observation
14-18
-------
ITEM#CC2
Topic- Establishment of Emergency Communication Systems
When/Where Applicable. Wherever a coordinated, large-scale response may become
necessary due to a major spill or discharge of a hazardous
material
Planning Goal. To quickly establish secure communication links between all
major parties to a response action
Action Items.
Designate the primary and alternate individuals responsible for communica-
tion links at important locations.
• Survey the kinds and types of communication systems used by various
agencies and organizations with a major role in emergency response,
particularly with respect to radio-communications equipment
Work out (at the very least) a system whereby the central command post
can communicate with all key parties to an incident and relay messages
between them regardless of the possible overloading of normal communica-
tion channels.
• Where necessary, have telephones with "unlisted" numbers installed at key
locations.
Where necessary, determine the procedure necessary to request the installa-
tion of new or additional telephone lines by the local telephone company at
various locations on an emergency basis.
Discussion-
Good communications are critical to a successfully orchestrated response action,
particularly if a number of different agencies or organizations at the local, county, state, and
even federal level have important roles Unfortunately, however, radio systems are unlikely
to be compatible among all parties, thus complicating overall coordination of activities.
Field personnel may not have ready access to telephones at all times. Incoming calls from
members of the public, news reporters, and a wide variety of others seeking information, can
tie up or slow available telephone service to emergency operations centers, police depart-
ments, fire departments, and offices of public officials, if not throughout the entire
community
14-19
-------
Where current radio systems are incompatible and there is limited opportunity to
replace or supplement existing units with compatible devices, the problem can be solved in a
fashion by planning to have each key responding agency or organization dispatch a
radio-equipped vehicle and driver to the central command post parking lot If the command
post itself has radios compatible with one or more of the mobile units, messages can be
relayed between various systems by drivers of the vehicles.
Incidents that are obviously going to require long-term response activities and/or
evacuations (i.e., more than a day or two) may necessitate the installation of temporary
telephone lines at various centers of activity (including the field command post) It is well to
know beforehand how to request such emergency service from the local telephone company
Alternatively, vehicles with cellular telephones may be stationed at the site
There are definite benefits for all emergency operation centers (including those at
major industrial facilities) to have several phone lines with unlisted and confidential
telephone numbers These can provide open lines of communications when all others are
tied up.
With respect to the overall use of telephones, take special precautions where an
explosion or fire at some critical location may destroy vital communication links or services
Give serious thought to organizing and taking advantage of the capabilities of local
ham radio operators in the area where and when their services may be necessary Many of
these individuals take their hobby very senously, are likely to have fixed station and portable
equipment that not only equals but exceeds the capabilities of local authorities (since most
ham radio operators truly enjoy attempting to contact distant states, countries, and even
continents with high power equipment), and will be more than willing to lend a helping hand
during an actual emergency.
It can also be prudent to establish a system for tracking, documenting, and prioritizing
messages, to test all vital communication links on a periodic basis, to inventory equipment
every so often, and to ensure vital equipment is properly maintained (and replaced with
alternate systems while undergoing repair).
14-20
-------
TTEM#CC3
Topic Formulation of Response Objectives and Strategy
When/Where Applicable- Wherever a spill, leak, fire, or potential explosion incident
involving hazardous materials may occur.
Planning Goal. To Establish procedures for formulating response objectives
and strategies, and to identify emergency situations in which
response personnel should limit their activities until the
situation has "stabilized" or until further information or expert
assistance has been obtained.
Action Items:
• Use the results of the hazard analysis to identify situations in which
response personnel should not intervene or should limit or restrict their
response actions.
• Establish procedures for evaluating hazards, risks, and site conditions so
that response objectives and strategy can be properly formulated at the
scene of an incident or accident.
Discussion:
Spill response guides published by industry and government alike often contain phrases
like "do not extinguish burning cargo unless the flow can be stopped safely." Additionally,
they almost universally advise response personnel to withdraw in the event of any sign that a
tank may explode (such as discoloration of the tank or rising sound from a venting pressure
relief device) or if the fire cannot be controlled by unmanned devices. More than one fire
department has planned to allow certain buildings (such as pesticide warehouses) to burn
while only protecting adjacent structures with water, thus preventing senous pollution
problems from chemically-contaminated runoff.
It follows from the above that there are potential accident scenarios in which the risks
associated with certain types of response activities may exceed any benefits to be realized,
thus providing ample reason to only undertake protective and containment actions from a
safe distance until the situation has stabilized or expert assistance has been obtained General
-Examples of situations in which the best course of action may be to hold back from a direct
"attack" include
• When a major release takes place that poses unknown hazards or hazards
for which response personnel are not equipped or prepared
• When a flammable gas or liquid is on fire and extinguishment could lead to l
release of toxic or flammable vapors or gases, and possibly, explosive
reigmtion /
14-21
-------
• When there are no endangered persons or structures nearby and the
container(s) and/or hazardous materials present significant hazards to
response personnel.
When the addition of water to a fire may serve to spread highly toxic
contaminants into the enviionment, thus causing a pollution problem that
may cost much more to resolve than the value of the burning materials,
vehicles, or buildings
Due to the possibility of such situations, response personnel should assess each
particular incident before taking action and formulate realistic response objectives. The
assessment should be based on:
• Pre-mcident plans and/or standard operating procedures
• Information that has been obtained regarding-
Matenal(s) involved
Container(s) involved
Vehicle(s) and/or structures involved
Atmospheric conditions affecting the incident
• Environmental monitoring and sampling data, if available
• Public protective actions that have or have not been initiated
• Resource requirements (i.e., trained personnel, specialized protective gear,
other equipment, etc.)
• Hazards and nsks posed to humans, animals, property, and the environ-
ment.
Upon completion of the incident assessment, command personnel will be in a better
position to determine whether their response strategy should be defensive or offensive in
nature. A defensive posture is best taken when intervention may not favorably affect the
outcome of the incident, will likely place emergency response personnel in significant
danger, and/or may possibly cause more harm than good An offensive posture (i e, one
requiring response personnel to work well within the boundaries of hazard zones) is best
taken when intervention is likely to result in a favorable outcome without exposing personnel
to undue danger and without causing new and potentially more severe problems In all cases,
of course, actions to protect the public and environment outside the immediate spill or
discharge area and/or to contain the hazard from a safe distance can be initiated regardless of
whether a defensive or offensive response strategy is chosen at the actual incident site
14-22
-------
ITEM#CC4
Topic: Ensuring Health and Safety at Incident Scenes
When/Where Applicable: Wherever a coordinated response may be necessary due to a
significant spill or discharge of hazardous materials
Planning Goal: To establish procedures for assuring the health and safety of
response personnel operating at hazardous material incidents
Action Items:
• Use the results of the hazard analysis to identify situations in which
response personnel may be exposed to chemical or physical hazards at an
incident scene.
• Establish standard operating procedures for a site safety and health program
that addresses:
Establishment of hazard control zones
Positioning of personnel, apparatus, and equipment
Selection and use of personal protective gear
Safe operating practices
Medical surveillance and care
Decontamination of protective gear and equipment
Decontamination of response personnel
Maintenance of field personnel
• Identify equipment and materials that may be needed to support the site
safety and health program.
• Establish procedures for obtaining any equipment and materials that may be
needed to support the site safety and health program
Discussion:
A system must be established to ensure the health and safety of emergency response
personnel The responsibility of establishing and managing the safety and health program
should be assigned to a predesignated Safety Officer and an alternate, who may also be
assigned one or more assistants
An important component of the safety program will involve establishment of hazard
control zones at the incident scene to limit the number of people in the most hazardous areas.
The exact size and configuration of these hazard control zones must be determined and
visually differentiated at each particular incident based on incident-specific factors and
situations and may include the following:
14-23
-------
• "Hot Zone" - Area of maximum hazard surrounding the damaged contain-
er(s) or fire area which may only be entered by specially equipped and
trained response personnel.
• "Warm Zone" - Area of moderate hazard outside the Hot Zone in which
properly equipped and trained backup crews standby and decontamination
takes place.
• "Cold Zone" - Area outside the Warm Zone that poses minimal or
negligible hazards to emergency personnel. The command post, most of
the deployed apparatus, and the resource staging area should be located in
the Cold Zone.
Safe operating procedures to be established and enforced by the Site Safety and Health
Officer include but are not limited to:
• The use of appropriate protective gear and equipment (see Items #PP1 and
#PP2 that follow).
• Limiting the number of personnel in the "Hot" and "Warm" hazard control
zones.
• Utilizing the most experienced personnel for the most hazardous tasks.
• Positioning a backup team in the "Warm Zone" in case it is needed to assist
or rescue personnel in the "Hot Zone".
• Providing medical surveillance for personnel before and after "Hot" and
"Warm" Zone operations.
• Monitoring (visually and through communications contact) the welfare of
personnel operating within the "Hot" and "Warm" Zones.
• Ensuring that all personnel understand their assignments.
• Ensuring that responders do not ingest contaminants through eating,
drinking, or smoking.
• Enforcing a "No Smoking" policy at incidents involving flammable or
combustible materials.
• Decontamination of protective gear and response equipment (see Item #PP3
that follows).
• Replacing fatigued personnel with "fresh" personnel.
• Adjusting hazard control zones to reflect changing conditions.
14-24
-------
Where deemed necessary, properly equipped medical personnel and one or more
ambulances should be available to check and (if necessary) treat injured or contaminated
response personnel as necessary. These personnel should check the vital signs and general
health of all personnel who will don specialized protective gear and enter "Hot" or "Warm"
Zones, particularly where fully encapsulating protective suits are being used, since
non-cooled suits can be very stressful to the wearer. The health of potentially exposed
response workers should be rechecked as appropriate and deemed necessary upon completion
of their duties
Readers should be advised that the Occupational Safety and Health Administration
(OSHA) has developed interim regulations for protection of workers at hazardous waste and
emergency spill sites. Covering all hazmat teams, local fire and police departments, and
emergency medical services, these regulations provide far more details than that provided
herein on how to go about meeting site safety and health requirements. Planning personnel
should therefore obtain the latest version of these requirements (see 29 CFR 1910.120) and
ensure they are prepared to comply with then* provisions Final rules on this topic are
expected during 1989.
The mtenm regulations currently in force reference several publications that provide
guidance on establishing safe operating procedures. A sample of these and others of
potential interest include:
Standard Operating Safety Guidelines, U.S EPA Office of Emergency
and Remedial Response, Hazardous Response Support Division, December
1984.
• The Decontamination of Response Personnel, Field Standard Operating
Procedures #7, available from same EPA office cited above, December
1984
• Preparation of a Site Safety Plan, Field Standard Operating Procedures #9,
available from same EPA office cited above, April 1985.
• Work Zones, Field Standard Operating Procedures #6, available from same
EPA office cited above.
• Personal Protective Equipment for Hazardous Materials Incidents^ A
Selection Guide, U S Department of Health and Human Services, Public
Health Service, Centers for Disease Control, National Institute for Occupa-
tional Safety and Health, October 1984 (see Items #PP1 and #PP2 also).
14-25
-------
ITEM#EV1
Topic: Designation of Decision Responsibility for Public Protective Actions
When/Where Applicable: Wherever public evacuations or other protective actions may
become necessary due to hazardous material spills or dis-
charges.
Planning Goal: To ensure the emergency plan denotes the specific mdividu-
al(s) with authority to initiate and later terminate an evacuation
or other protective action under varying circumstances.
Action Items:
• Check local and state laws and regulations to determine who has responsi-
bility for evacuation or "shelter-in-place" decisions.
• Ensure that this individual is aware of the responsibility and prepared to act
in a timely fashion
Discussion:
Laws designating authority for public evacuations, activation of the Emergency
Broadcast System (EBS), or activation of other means to alert or notify the public of an
emergency vary from state to state and even possibly at the local level Since time is critical
during the initial stages of a real or potential emergency, it is necessary to know who has
authority to make evacuation and similar decisions. In addition, it is necessary to ensure that
this (these) individual(s) will be contacted promptly, will be given the information needed to
make proper decisions, and will be capable of acting quickly and decisively
Where these decisions are the responsibility of elected officials such as the governor of
the state or local mayors, it may be prudent to ask these individuals to delegate the
responsibility to the person(s) given overall command of response activities, particularly if
these latter personnel are better qualified to assess the nature and magnitude of the threat
Once the emergency is over, a person in authority must give approval to permit reentry
of evacuated areas or to give an "all clear" signal The response plan should also identify this
individual.
14-26
-------
ITEM#EV2
Topic: Criteria for Evacuation and Shelter-in-Place Decisions
When/Where Applicable* Wherever public evacuations or other protective actions may
become necessary due to hazardous material spills or dis-
charges.
Planning Goal. To give command personnel guidelines with respect to* 1) the
circumstances under which the public should be evacuated or
instructed to shelter-in-place; and 2) the size and shape of the
areas that should be considered vulnerable during various
types of emergency situations
Action Items.
Make policy decisions with respect to levels of toxic agents, thermal
radiation, and explosion overpressures which may be tolerable by the
public.
• Compile data on the size and shape of potential hazard zones.
• Develop criteria for evacuation and shelter-in-place decisions.
Discussion*
This is not a topic commonly suggested for inclusion in hazardous materials emergency
response plans but is one that definitely warrants attention Time can be critical in situations
that may require public evacuations or related protective actions and may not permit lengthy
discussions or deliberations as to whether an evacuation is warranted or how large an area
should be considered at risk. Predetermined criteria for decision-making can therefore
greatly facilitate the process
The accident scenarios developed via use of this guide, as well as the computational
methods described in Chapter 12, can be a key source of information for evacuation planning
where specific facilities or hazardous material shipments are known to pose a threat These
threats can be quantified via use of this guide and tabulated within the emergency response
plan. Separate tabulations can be prepared for toxic vapor clouds or plumes, liquid pool
fires, flame jets, potential BLEVEs, and the other specific hazards addressed by Chapter 12
Where alternative guidance is lacking, and the incident scenario is not one that was
considered during the planning process, some thought may be given to the suggestion that the
methods of Chapter 12 be applied on a real-time basis during emergencies to evaluate hazard
zones
14-27
-------
The 1987 Emergency Response Guidebook (DOT P 5800 4) contains guidance from
the U.S. Department of Transportation (DOT) with respect to recommended evacuation
distances for potential BLEVEs and vapor or gas hazards involving rail or highway vehicle
accidents. The list of hazardous materials considered for gas or vapor hazard purposes (in an
appendix to the DOT guide) is relatively short but includes those highly volatile substances
most commonly transported in commerce.
This same section of the emergency plan should also give some guidance as to whether
the public should be evacuated from a toxic vapor hazard zone or told to shelter-m-place ~ or
whether both protective measures should be considered for use m different portions of the
zone (i.e., evacuation close in to the vapor source where concentrations are higher,
shelter-in-place at further distances) See Appendix C to this document for further informa-
tion on evaluation and implementation of the shelter-m-place option
In all cases, remember that hazardous material spills can be "dynamic" events in the
sense that incident conditions, the weather, and the wind direction can change with time.
Guidance obtained from consequence analysis procedures provided in this document or other
sources should only be considered a starting point for the decision process In an actual
emergency, evacuation area and/or hazard zone assessment must be a continuous activity.
14-28
-------
TTEM#EV3
Topic: Public Alert/Notification/Instruction Procedures
When/Where Applicable- Wherever public evacuations or shelter-in-place decisions may
become necessary due to hazardous materials spills or dis-
charges.
Planning Goal: To plan for and specify the methods and procedures by which
the public will be alerted of an emergency situation and given
instructions on what to do.
Action Items:
• Identify available alert and notification methods and practices.
• Specify which are to be used, who will use them, and when and how they
are to be activated.
• Have lists of suitable shelter facilities available and procedures specified
for assigning different groups of people to these various locations.
• Specify evacuation routes for all areas at risk based on results of the hazard
analysis.
• Prepare sample messages for various situations.
• Establish a procedure to obtain and distribute appropriate protective
equipment to personnel who may experience prolonged or excessive
exposures to toxic contaminants while performing notification or evacua-
tion operations. Provide training in use of this equipment as necessary.
Discussion:
Public authorities require the means to alert the public of an emergency situation and to
give instructions to those individuals within a hazard zone or approaching such a zone.
Options for alerting the public include community or industrial facility horns or sirens,
use of the Emergency Broadcast System (BBS), broadcasts by individual radio and television
stations (including cable TV), use of police or fire department vehicles with public address
systems, door-to-door coverage of neighborhoods by knocking on doors, use of an "all-call"
system which rings all telephones in the area and repeats a recorded message, use of
helicopters with public address systems, and various combinations of the above. Be advised
that the public must be instructed before an actual emergency as to the meaning of warnings
provided by horns or sirens via a public education program. Emergency planning requires
knowledge of the specific procedures and access codes for utilization of the EBS and radio
and television station resources. As an evacuation progresses, police, fire, public works,
14-29
-------
and/or other government employees (depending on who might be the most readily available
and tree of other important duties at the time) may have to go door-to-door to ensure that all
residents have been alerted and to provide assistance to the elderly, physically handicapped
or hearing unpaired It is a good idea for these personnel to be equipped with appropriate
personal protective equipment when necessary and a supply of chalk or colorful tags that can
be used respectively to mark doorsteps or place on door knobs to indicate that the building
has indeed been evacuated. Note that evacuation of people in individual residences who
require special notification or assistance can be facilitated if public officials have compiled a
list of those homes requiring special attention
In designating evacuation routes, keep in mind that major roads are most desirable, but
may not always be available. Since the direction of the wind at the time of a hazardous gas
or vapor release cannot be predetermined, and since the direction may change with time,
emergency personnel may require more than one option for any given hazard zone. As soon
as an evacuation has been declared, police and auxiliary personnel should be prepared to
control traffic on evacuation routes, to keep non-evacuation related traffic off these roads,
and to remove any vehicles that breakdown and cause a slowdown of movement. These
activities may in turn require the ready availability of tow trucks and portable roadblock
materials (barricades, cones, signs, etc) Additionally, thought should be given to removing
impediments to traffic flow caused by excessive precipitation (rain or snow), fallen trees,
crossing trains, and so forth.
Standard message formats for use in radio and television broadcasts can facilitate and
reduce the time necessary to alert the public of a problem and inform them of the protective
actions to be taken. The overall planning process must consider designation of evacuation
routes and these routes should be identified in public warning messages, as should the
location of shelters to which people with automobiles should proceed, the locations where
people without automobiles should gather for pick up by buses or vans (see item #EV4), and
what actions should be taken by people with children at school in a potential hazard zone
These messages should instruct people to bring any prescription medicines and special
personal care items with them See item #EV7 for a discussion of what to do about pets.
The Hazardous Materials Contingency Planning Course manuals referenced in Chapter 1
provide several sample messages for consideration in accomplishing this objective. Keep m
mind that messages may have to be broadcast in languages other than English in
communities with concentrations of ethnic minorities
Incidentally, be advised that there are likely to be people who for one reason or another
may resist instructions to evacuate, even when confronted face-to-face with a police officer
Officers sent to persuade the last few families or individuals who refused to evacuate in a
major 1986 evacuation in Miamisburg, Ohio, achieved considerable success by asking these
people to provide information on their "next of km" for use in the event of their demise. The
holdouts, to the last person, got the message and cooperated with the evacuation order
14-30
-------
ITEM#EV4
Topic. Transportation Functions
When/Where Applicable* Wherever public evacuations may become necessary due to
hazardous material spills or discharges and public authorities
find it necessary to provide transportation for evacuees
Planning Goal(s): To plan for the availability of buses, vans, ambulances, and
other vehicles to transport general members of the public
without automobiles, school children, residents of hospitals
and nursing homes, and possibly even pnson populations to
safe shelters
Action Items*
Assign one or more persons responsibility for coordination of transporta-
tion activities during an emergency.
Establish agreements with public and private bus companies and ambulance
services for provision of vehicles and drivers under emergency conditions
Develop evacuation plans for special occupancies such as schools, day care
centers, hospitals, nursing homes, and prisons and the like, possibly
coordinating efforts with similar facilities in some sort of mutual aid
program
Develop communications, dispatch, and command/coordination systems for
control of the vehicle fleet.
Bnef drivers on procedures periodically.
Establish a procedure to obtain and distribute appropriate protective
equipment to personnel who may expenence prolonged or excessive
exposures to toxic contaminants while performing notification or evacua-
tion operations.
Consider giving training to at least a portion of the drivers in the use of
self-contained breathing apparatus (SCBA) if entry might be necessary or
may unexpectedly occur (due to wind direction shifts or other factors) to
zones with toxic air contamination
Select safe locations (to the extent possible) at which the public should
assemble in then: respective neighborhoods for pick-up by transport
vehicles
14-31
-------
• Where large areas may be at nsk (based upon results of the hazard
analysis), plan for a staged evacuation with zones at highest nsk being first
evacuated, followed by zones with lesser nsks.
Discussion:
The action items are self-explanatory. Note that a good evacuation plan is not only
useful for hazardous material emergencies but any other emergency situation that may
require relocation of the public The presence of hospitals, large schools, nursing homes,
and/or prison facilities can require detailed planning and preparedness to facilitate what will
be a monumental task under the best of circumstances
The individual given responsibility for transportation operations might also be given
responsibility for ensuring that any needed response equipment, materials, and personnel are
delivered promptly to the scene of an accident and for ensuring an adequate state of
operational readiness, thus consolidating the management of all transportation related
activities.
14-32
-------
ITEM#EV5
Topic: Care and Shelter of Evacuees
When/Where Applicable: Where public authorities may find it necessary to provide
shelter for large numbers of evacuees
Planning Goal(s)' To provide safe and comfortable shelters for relocated
populations.
Action Items.
• Use hazard analysis results to estimate the maximum number of people for
which shelter might need to be provided and the possible duration of a
major evacuation, taking into account those families and individuals that
might stay with friends or relatives or prefer to check into hotels and
motels.
• Coordinate shelter planning efforts with the American Red Cross and
ensure that the local chapter can cope with the number of potential
evacuees.
• Have lists of shelters and route maps readily available for use in giving
instructions to evacuees and vehicle drivers.
• Where necessary, ensure plans are in place for populations needing special
care.
Discussion.
High schools with showers and cafeteria facilities are probably the best choices for use
as temporary shelters, with large churches having function halls a viable alternative in most
communities During vacation seasons, universities or other schools with dormitory facilities
should be considered Not to be forgotten are any nearby military bases with excess or
temporary housing facilities and overnight summer camps for youngsters in the off-season
The American Red Cross has long established methods and procedures for sheltering
evacuees Determine its capabilities in the locale or jurisdiction of concern, consider its
needs in overall emergency planning, and plan to provide any assistance required to achieve
the stated planning goal Helpful hint Work with the local chapter in developing evacuation
instructions to be broadcast to the public, particularly with respect to the clothing, bedding,
medicines, and other supplies the Red Cross may wish the public to bring with them
Once evacuees reach a shelter, people will want to report "missing" persons or to
determine if their friends, family members, relatives, or neighbors are "lost" or at another
shelter Response to these quenes, as well as identification of persons legitimately missing,
will require registration of people upon entry and communications between shelters Where
14-33
-------
the evacuation may be prolonged, and vehicles are available for use after the primary
evacuation has been completed, a mechanism should be established for reunion of separated
families, relatives, and close friends that can provide a mutual support system under adverse
circumstances.
Shelters should be assigned medical teams to care for people who become ill during the
evacuation or at later times. These medical personnel should be alerted to the signs and
symptoms of exposure to the hazardous matenal(s) involved in the incident so that they may
identify victims and provide necessary care. Contaminated individuals (those having
contaminant residues on their persons or clothing) should be segregated from unexposed
people until decontaminated. (Note: Significant contamination is unlikely to be of concern
except where highly toxic aerosols, mists, or dusts have entered the atmosphere or where
individuals were in the immediate vicinity of the spill or discharge) Facilities should also be
available for care of the handicapped or elderly.
Thought should be given as to how best to manage any pets brought along by evacuees
Human services personnel may be necessary to fulfill counseling, recreational, and other
needs of confined populations. Quite obviously, shelters will require initial and periodic
supplies of food, water, and all other personal need items of inhabitants
14-34
-------
ITEM#EV6
Topic Security of Evacuated Hazard Zones
When/Where Applicable: Wherever large areas may require evacuation due to hazardous
material spills or discharges.
Planning Goal: To ensure that private and public property is safe-guarded
during evacuations and that unauthorized individuals do not
enter hazard zones
Action Items:
• Plan to provide police and/or other forces to man roadblocks on routes to
potential hazard zones.
• Plan for ground and possibly aerial patrols of evacuated areas.
• Establish procedures to provide personnel with appropriate personal
protective devices and training hi their use where changing accident or
environmental conditions may suddenly change the boundaries of haz-
ardous areas.
Discussion
Personnel and vehicles should be available to establish and maintain roadblocks on the
boundaries of hazardous areas. One reason for this is to prevent uninformed or curious
members of the public from entering hazardous locations Another involves prevention of
entry by those who may wish to take unlawful advantage of the fact that some part of the
community has been temporarily abandoned
Where and when it is safe to do so and deemed necessary, patrols of evacuated areas
can enhance security while double-checking that all members of the public have left
designated areas Aenal patrols during daylight hours can provide a better "picture" of what
is happening on the ground, while helicopters can help ferry critical personnel and supplies
Helpful hint: Helicopters may be available from the state police, local charter services, local
radio and television stations, and nearby military posts.
14-35
-------
ITEM#EV7
Topic: Movement and Care of Livestock and Pets
When/Where Applicable: Wherever evacuations may become necessary due to haz-
ardous material spills or discharges
Planning Goal: To plan for and specify the methods and procedures by which
domestic or other captive livestock may be moved to a safe
location and properly cared for when the potential arises for a
major toxic gas or vapor release to the environment and tome
permits this action. Also, to give some thought to tending pets
left behind in an evacuation or brought to shelters by evacuees
Action Items'
• Use the results of the planning basis development process to determine if a
large number of valuable livestock might be threatened by a major spill or
discharge.
• Decide whether resources and time might be available for this activity or
whether it is only practical to protect human populations
• If desired and necessary, plan for the movement and care of livestock with
local resources
• Establish procedures for handling household pets during evacuations
Discussion:
Movement and care of domestic livestock is likely to be most applicable in rural areas
with large populations of valuable animals and relatively few people It may not be worthy
of consideration in other circumstances unless a zoo containing extremely valuable or rare
animals may be at risk.
The subject of household pets may seem rather trivial at first, but it is well to remember
that many people care deeply for their animal friends, and the issue of pets has caused
problems in past accidents Planning personnel will have to decide whether to permit
evacuees to bring their pets with them to shelters or to mandate that they be left behind, with
the knowledge that both options are surely to cause difficulties of one kind or another
As time passes during an evacuation in which pets have been left behind, and the
evacuation was ordered because of the threat of a release rather than an actual discharge,
people will ask questions about what is being done to feed then: animals and/or may attempt
to enter evacuated areas to care for them One way to handle the problem for pets left
outdoors is to assign someone the responsibility of leaving supplies of water and pet foods at
various locations on a daily basis when and where it is safe to do so People who are forced
14-36
-------
to leave pets indoors can be told to set out several days of food and water before leaving
home Keep in mind that pets exposed to toxic agents may be injured or killed and that
hazard zone reentry activities after the threat has abated should include procedures to collect,
care for, and possibly dispose of these animals as necessary and appropriate.
Even if evacuees are told to leave pets behind and not bring them to shelters,
emergency preparedness personnel should expect and plan for the fact that some people will
indeed bring then: pets with them and should have procedures worked out on how to handle
these situations.
14-37
-------
ITEMtfFFl
Topic: Special Firefighting Equipment and Materials
When/Where Applicable: Any locale in which a fire involving hazardous materials may
require special equipment or extinguishing agents for control
and/or extinguishment
Planning Goal: To ensure that the fire service has rapid access to any special
equipment and materials it may need in the event of an unusual
fire situation
Action Items:
• Use the results of the hazard analysis to identify those sites or spill
scenarios that appear to present unusual fire hazards in the sense that
specialized equipment and/or supplies may be required for response
• Identify sources of needed firefighting equipment or supplies that may be
called upon in an emergency.
Discussion:
Public fire departments primarily rely upon water for fire control and extinguishment,
but are well aware that water can be ineffective on some types of fires and may actually be
counterproductive or dangerous for use in other cases Nevertheless, since water is usually
adequate for most large fires encountered, these departments rarely stock more than a limited
number of portable dry chemical or carbon dioxide extinguishers and possibly a small
amount of foam concentrate These supplies of auxiliary agents may not be adequate for
major chemical or petroleum product fires or fires involving combustible metals Thus,
where the hazard analysis has identified scenarios requiring unavailable resources, the fire
department should be assisted in identifying sources of additional supplies and equipment for
use in emergencies.
Airports comprise one potential source of assistance Military airports are likely to
have crash/rescue vehicles with significant foam generation and dispensing capabilities
They are also likely to have large capacity units for discharging carbon dioxide, halons, and
possibly dry chemicals Civilian airports are likely to have large foam trucks and/or dry
chemical trucks, and wheeled-portable dry chemical units Both types of airports may have
supplies of special agents for combustible metal fires
Chemical and petroleum processing facilities with internal fire brigades are another
potential source of supplies They may also have special portable equipment for fighting
storage tank fires.
14-38
-------
ITEM#FF2
Topic* Identification of Water Sources in Rural Areas
When/Where Applicable. Where an accident requiring large amounts of water for
response takes place in an area distant from hydrant or other
water supply systems
Planning Goal' To identify sources of water in areas not served by a central
water supply system
Action Items:
• Use hazard analysis results to identify potential incident locations distant
from water systems
• Compile a list of rivers, streams, lakes, ponds, reservoirs, wells, farm
holding tanks, other water tanks, and swimming pools that can be used to
supply water to an accident scene or to refill water trucks.
• Establish procedures for coordinating and implementing water supply
operations
Discussion:
The planning goal and action items are self-explanatory.
14-39
-------
ITEM#HC1
Topic: Establishment of Special Rescue Squads
When/Where Applicable. Wherever response personnel or members of the public may be
disabled or trapped due to a spillage, fire, or explosion
involving hazardous materials.
Planning Goal: To have two or more special teams available who can safely
enter hazardous environments to rescue injured or trapped
individuals
Action Items:
• Assign responsibility for special rescue operations to specific teams
• Use the planning basis scenarios to identify potential situations requiring
special clothing or equipment for safe rescue operations.
• Ensure that trained rescue personnel have or can quickly obtain self-con-
tained breathing apparatus (SCBA) and any required chemical protective
clothing.
• Ensure that spare breathing apparatus is available for use by those being
rescued where necessary.
Discussion:
There are a variety of scenarios under which workers at chemical facilities or members
of the public near or downwind of a hazardous material release may be overcome by toxic
vapors or gases, exposed to high levels of thermal radiation, or injured due to the effects of
an explosion. Fire departments are usually well-prepared and experienced in rescuing people
from fire and explosion situations, and will in many cases not require any new or additional
planning to meet these responsibilities. The situation can be quite different, however, where
toxic or corrosive chemicals may have been released to the environment or continue to be
released.
Chapter 6 explained that some chemicals can be readily absorbed through the skin to
cause toxic effects and that others can have a corrosive action on bodily tissues A problem
arises when: 1) such materials are on the ground, must be walked through to reach victims,
and are incompatible with the usual footwear of rescue personnel, thus possibly allowing
contamination of the feet; or 2) high concentrations of such substances in air can penetrate
the normal clothing of rescue personnel. In either case, rescue workers may need special
chemical protective clothing, together with a self-contained breathing apparatus (SCBA), to
carry out their mission without themselves falling victim. There are benefits, therefore, in
14-40
-------
assigning these unusual rescue missions to special teams that are trained and properly
equipped for the duty. More on the topic can be found in items relating to personal
protection.
The purpose of suggesting the availability of spare SCBA units is actually two-fold
Not only might people trapped m hazardous area require them to escape, rescue workers may
need extra air supplies to accomplish lengthy rescues Even the best SCBA units rarely
provide air to the wearer for more than 30-60 minutes Heavy exertion while wearing these
units can significantly shorten then* duration of effective operation
Rescue teams operating in hazardous environments should work in at least pairs. This
is a common safety practice, as is the practice of having a backup team ready for action if a
problem should develop.
Rescue teams of a different nature and with different equipment requirements may be
necessary when buildings have collapsed due to an explosion and people are trapped under
the rubble.
14-41
-------
ITEM#HC2
Topic. Provision of Ambulance Services for Victims
When/Where Applicable: Wherever the number of casualties due to a hazardous material
spill or discharge may exceed the normal capacity of available
ambulance services
Planning Goal: To ensure that sufficient ambulances and EMTs are available
for mass casualty situations (in addition to units that may be
needed for evacuation of health care facilities).
Action Items:
• Use hazard analysis results to determine worst case ambulance service
needs.
• Establish mutual aid systems with services in neighboring communities and
regions as needed.
• Determine procedures for requesting military medical assistance if major
bases are nearby.
• Establish procedures that may be required by EMTs m treatment and
decontamination of victims exposed to chemical contaminants, including
the administration of special antidotes that may be necessary in the event of
poisoning by certain highly toxic substances
• Establish procedures for obtaining and distributing potentially required
personal protective clothing and devices to ambulance personnel, provide
necessary training for their use.
• Ensure that any decontamination and/or treatment supplies required by
ambulance personnel will be readily available in an actual emergency.
Discussion:
The action items are self-explanatory. While undertaking planning activities, keep m
mind that populations normally residing outside the established hazard zone will continue to
require access to emergency ambulance services. With respect to the potential problem
involving contamination of victims and vehicles by chemical residues, be advised that this
will usually only be of concern for relatively few hazardous materials released into the
atmosphere as aerosols or dusts (when the victims have not been in the immediate vicinity of
the spill or discharge). The vast majority of airborne gases or vapor will not "stick" to a
person or his/her clothing in any significant fashion See Item #HC3 for contamination
problems in health care facilities and Item #PP3 for decontamination of protective clothing
and other response equipment for additional information
14-42
-------
ITEM#HC3
Topic: Establishment of a Mass Casualty Plan
When/Where Applicable' Wherever mass public casualties are likely in the event of a
disaster.
Planning Goal: To ensure that area hospitals and health care facilities are
prepared to handle mass casualty situations, including those
due to spills or discharges of hazardous materials.
Action Items:
• Use the hazard analysis results to determine which hazardous materials
accidents could cause mass public casualties.
• Ensure that area hospitals and health care facilities have a coordinated plan
for handling mass casualty situations in general.
• Ensure that necessary advice for medical treatment of chemical exposures
is readily available.
• Ensure that means exist to rapidly obtain any special antidotes or medical
supplies that might be needed on short notice and in large quantities.
• Work with the American Red Cross, the Salvation Army, and members of
the local clergy in developing procedures for notification of next of kin in
the event of fatalities or senous injuries.
Discussion:
This item pertains to those emergency situations which have the potential to kill or
injure dozens, hundreds, or possibly thousands of individuals over a short penod of time. Its
intent is to ensure that medical care providers can cope with the problem as efficiently and
effectively as possible. Fortunately, many hospitals and clinics across the country already
have such plans for non-chemical related emergencies and therefore have the basic elements
of a plan that can be expanded to cope with hazardous material emergencies. There are,
however, five special topics to consider.
First of all, it is necessary in all chemical exposure situations, not just those involving
large numbers of people, to have information readily available on: 1) the toxic effects of the
substance(s) of concern by all likely routes of exposure; 2) the observable symptoms of
human exposures, 3) the special medical tests (if any) that may be advisable to assess the
extent of injury; 4) the need to observe victims for delayed effects; and 5) the treatment
methods or protocols recommended for various types and levels of exposure. Public
authorities should never assume that physicians or hospitals have this information on hand.
14-43
-------
Nor should they assume that the basic first aid and health effect data given in typical material
safety data sheets will be adequate for all eventualities Options to identify or obtain more
comprehensive information include:
1) Compile or arrange rapid access to a library of detailed chemical toxicology
guides and handbooks.
2) Arrange access to computerized on-line toxicology data banks such as the
major system operated by the National Library of Medicine
3) Rely on GHEMTREC to identify appropriate information sources at
chemical manufacturing companies in the aftermath of a spill or discharge
4) Request that industrial complexes handling hazardous materials prepare a
- medical response plan for chemical emergencies and appoint one or more
liaisons to area medical facilities, preferably selected from their internal
medical department staffs.
5) Make contact with local Poison Control Centers to determine their
capabilities.
Mass casualty situations may require establishment of field hospitals to care for the
injured and to identify, stabilize, and transport more serious cases to hospitals. Some
consideration should be given to preselection of sites to which the public can be directed in
the aftermath of a spill emergency where this action is warranted. Outside and local medical
care personnel should be informed of their responsibilities in staffing and equipping such
facilities on a rapid basis. Emergency plans should consider the need for accurate accounting
of patients and their destinations after triage. The onscene medical command post should be
under the direction of a single individual and an alternate in charge of all medical operations.
A supplement to the Guide fot Development of State and Local Emergency Operations
Plans referenced in Chapter 1 of this guide and dated March 18, 1987, provides extensive
guidance on planning for this contingency and is highly recommended. FEMA's document
designation for this supplement is CPG1-8, CHG1.
Some chemical exposures that can poison the body are best treated with specific
antidotes or special equipment. Medical response plans should identify and ensure availabil-
ity of any such special supplies. This does not necessarily mean that large stocks must be
purchased and stockpiled, only that sources be identified and a mechanism be established for
their rapid procurement Helpful hint: Ask the medical departments of local industries
handling toxic chemicals whether they stock or see a need to stock special supplies or
antidotes for chemical exposures.
Be aware that victims of a chemical-related emergency may be contaminated in the
sense that their skin, hair, or clothes may have residues of chemicals under several
envisionable accident scenarios, particularly if they have been in the immediate vicinity of
the spill or discharge or exposed to airborne aerosols or dusts. There have been cases where
14-44
-------
medical personnel have had to decontaminate victims and have needed protective clothing (at
least gloves and masks) to protect themselves from potential toxic exposures Medical
response plans should take such possibilities into account.
Finally, besides the possible need for a temporary morgue and large-scale mortuary
services, note that next of km must be promptly notified of fatalities or severe injuries
carefully and in a sensitive and supportive fashion. As noted elsewhere, this activity should
be discussed with the American Red Cross and Salvation Army and coordinated with
members of the local clergy.
14-45
-------
TTEMtfPPl
Topic: Availability of Respiratory Protective Devices
When/Where Applicable: Any locale that may expenence a major release of irritating or
toxic airborne contaminants
Planning Goal: To ensure both initial and continued availability of a sufficient
number of self-contained breathing supply apparatus (SCBA)
during emergency situations
Action Items:
• Inventory the number and types of units currently available to local fire
departments
• Determine the circumstances under which available equipment can be used
safely and ensure that all limitations on use are clearly understood.
• Determine local capabilities for refilling of air supply tanks
• Attempt to estimate the need for such units under worst case conditions
determined during the hazard analysis process
• If a shortage may develop, identify sources of additional equipment and
refilling capabilities that may be called upon in an emergency
Discussion:
Fire departments will in most cases have a sufficient number of self-contained
breathing supply units and refilling systems to meet the initial needs of then: own personnel
in chemical related emergencies There may be situations involving major discharges of
airborne contaminants, however, where an insufficient number of units are readily available
to meet the potential needs of police officers, ambulance personnel, spill containment and
cleanup personnel, public authorities and officials at the scene of the incident, and others who
may need to function in an are? that could be suddenly exposed to hazardous vapors, gases,
or aerosols in the event a container fails or the wind shifts direction It is therefore important
that this topic be given special attention during the overall emergency response planning
process.
The need for training in the use of these devices is addressed m Item #TR1. Keep in
mind that it is difficult or impossible to wear a self-contained breathing unit while driving or
riding in many types of vehicles This should be given consideration during the planning
process.
14-46
-------
Expert guidance on selection and use of personal respiratory protective devices can be
found in the N1OSH Guide to Industrial Respiratory Protection, available from Publications
Dissemination, DSDTT, National Institute for Occupational Safety and Health, 4676
Columbia Parkway, Cincinnati, Ohio, 45226 (telephone 513-841-4287). Additional informa-
tion may be found in publications cited in Item #PP2.
14-47
-------
Topic: Availability of Special Protective Clothing
When/Where Applicable. Any time that rescue, leak plugging, spill containment, or
other response activities may directly expose individuals to
sudden fires or corrosive, irritating, or toxic solids, liquids,
gases, vapors or fumes that may harm the individual via direct
contact with the skin or eyes.
Planning Goal: To ensure the ready availability of any special chemical or
thermal protective clothing that might be needed by response
forces, particularly during the initial phases of a response
action.
Action Items:
• Use the results of the hazard analysis to determine the spill scenarios and
related chemicals that may require local forces to enter highly contaminated
and/or fire prone environments.
• Determine the types of protective clothing that may be required to permit
response workers to enter these environments.
• Determine which clothing materials are appropriate for the expected
exposures.
• Arrange for rapid availability of appropriate protective clothing in the event
of an emergency.
Discussion:
The normal turnout clothing of fire service personnel may be fully adequate to protect
these individuals in a wide variety of fire and/or spill situations. But there may also be cases
where rescue teams or individuals who desire to enter the immediate spill area for leak
plugging or spill containment purposes might be exposed to corrosive substances or toxic
substances that might be absorbed through the skin Such situations may require more
complete protection of the body by clothing that is resistant to the damaging effects of the
spilled substance. The clothing itself may range from boots, gloves, or disposable suits made
of chemical resistant materials to air-tight fully encapsulating "astronaut" suits that offer
complete protection of the body from spilled substances on the ground or in high
concentrations in the air. Where fires may occur, appropriate thermal protection may be
additionally necessary.
14-48
-------
The first action item takes advantage of the information gathered during the hazard
analysis process to identify the hazardous substances and related spill scenarios that may
require the availability of specialized clothing The second action item indicates that
emergency planners should determine the types of specialized clothing that may be necessary
(i e., gloves, boots, aprons, face shields, splash suits, astronaut-type suits, etc), while the
third notes it is necessary to ensure that the materials from which the clothing is constructed
will not be penetrated by the spilled substance Finally, the last item suggests that a
mechanism be established to ensure that the necessary clothing is readily available when
needed
Several guides to the selection of protective clothing for spill response are.
Guidelines for the Selection of Chemical Protective Clothing, sponsored
by the EPA and available from the ACGIH Publications Section, 6500
Glenway Avenue, Bldg. D-7, Cincinnati, Ohio 45211, or (513) 661-7881.
• Occupational Safety and Health Guidance Manual for Hazardous Waste
Site Activities, prepared jointly by NIOSH, USCG, and EPA and available
as DHHS (NIOSH) Publication No 85-115 from the U.S Government
Printing Office
Personal Protective Equipment for Hazardous Material Incidents: A
Selection Guide, National Institute for Occupational Safety and Health,
DHHS (NIOSH) Publication No. 84-114, available as document PB
85-222-230 from the National Technical Information Service, Springfield,
VA 22161.
Note that it may not be cost effective for every community to purchase every possible
type of protective clothing that may become necessary There are advantages to cooperative
agreements on a county or regional basis (with industry or government institutions) that
permit individual communities or firms to draw upon a central stockpile. There are also
advantages to giving consideration to the likelihood that an accident requiring certain
clothing will actually take place in the foreseeable future. It does not necessarily make sense
to purchase something with a shelf-life of 5-10 years at high cost if probability analysis
procedures have indicated a potential need once every 200 years or so on average.
14-49
-------
ITEM#PP3
Topic: Decontamination of Exposed Protective Clothing and Other Response Equip-
ment.
When/Where Applicable: Wherever a spill situation may contaminate the protective
clothing of response personnel or other items of equipment.
Planning Goal: To ensure that response personnel are not contaminated with
corrosive or toxic materials while removing protective clothing
worn in a contaminated environment; similarly, to prevent
injury of people later using items of contaminated equipment
Action Items:
• Establish decontamination procedures for clothing and equipment in the
emergency response plan.
• Assign responsibility for clothing and equipment decontamination to a
qualified individual.
Discussion:
Clothing and equipment used in a contaminated environment may itself become
contaminated. Clothing must be decontaminated before it can be safely removed by its
wearer. Contaminated equipment may need careful decontamination before being safe to
touch or use in the aftermath of an incident.
In many instances, it will be sufficient to merely wash the clothing or equipment down
with strong water sprays or large amounts of water Several manufacturers market decon-
tamination showers for wearers of protective clothing, these consisting of a framework of
water piping with numerous water spray nozzles surrounding an open space the size of a
shower stall. Fire hoses and possibly even garden hoses can be used to wash down most
equipment as well.
There are two potential complications to be considered First of all, it must be realized
that the water and/or other solutions used to decontaminate the clothing or equipment may
contain some amount of the contaminant It must be decided on a case-by-case basis whether
this water should be contained, collected, treated, and/or sent to an appropriate wastewater
disposal facility, or whether it can simply be released to the environment. Secondly, it must
be realized that not all contaminants may be completely washed off by water alone Final
decontamination may require the careful use of various solvents or cleaning compounds
These may range from ordinary soap to specialized chemicals designed to neutralize
remaining residues
14-50
-------
ITEM#PR1
Topic: Public Relations in Emergency Situations
When/Where Applicable Any locale subject to a significant hazardous material incident.
Planning Goal Preparations for keeping the public informed of continuing
developments via newspaper and broadcast media while
minimizing rumors, conflicting reports, confusion, and unnec-
essary interference with the activities of key response person-
nel -- and while also coordinating media relation efforts with
those of other agencies or parties to the response
Action Items.
• Designate one specific individual and an alternate press officer to join the
team of press officers that may be formed from representatives of all major
parties to an emergency response operation (be they from local, state or
federal government agencies, or the company responsible for the accident).
• Compile a list of telephone numbers of local radio and television station
personnel who can initiate special "on air" announcements
• Provide designated press officers with secretarial support, photocopy
machines, and a means of communications with the overall commander of
the response operation
• Select a site, preferably but not necessarily near the central emergency
operations center, where the press can convene and be briefed by the press
officer team This site should ideally have telephones, electrical outlets,
restrooms, and other facilities that media personnel may require
• Establish a firm policy among all local officials and response personnel as
to who should or should not speak to media personnel
• Ensure that key emergency response personnel understand the need to relay
up-to-date "status reports" to press officers on a regular basis
Discussion
The public needs to be informed accurately and rapidly as to what is happening during
an emergency situation Significant incidents may result in a large number of reporters
arriving on scene and attempting to interview anybody and everybody The above action
items will help to reduce confusion, facilitate information transfer, reduce problems that
might be otherwise caused by a lack of organization, and reduce the incidence of unfounded
rumors
14-51
-------
The book by AJ. Smith, Jr., referenced at the beginning of this chapter, has an
excellent section addressed to public officials on the subject of dealing with the media which
discusses many "do's" and "don'ts" These mostly pertain to the actions of press officers for
government agencies, and include:
THINGS TO DO
1. Accommodate the media as much as possible, make the news available to
them.
2. Schedule news conferences and avoid written releases.
3. Be direct and specific.
4. Always, always tell the truth, don't hedge.
5. Hold public hearings at least twice during a week-long event and invite the
press.
6. Have news conferences immediately after any meeting from which the
media or public have been barred.
7. Send a press representative to the command post
8. Ensure that the team of press officers is in contact with the command post
at all times.
9. If safety permits, allow the media to take pictures of the accident site
THINGS NOT TO DO
1. Do not permit arguments among public officials or press officers from
different organizations in front of the press. Do, however, permit statements
of dissenting opinions
2. Avoid giving gut opinions or conjecturing
3. Do not be evasive If the answer to a question is not known, refer the
question to someone who has the appropriate answer
4. Do not be critical in a personal manner; i e, avoid personal remarks about
other people at the accident scene.
5. Do not be philosophical These kinds of discussions are extremely suscep-
tible to being quoted out of context
14-52
-------
6. Do not make off-the-record comments. They may end up in print with later
retractions buned in the back pages.
7. Avoid friendly chats with media people. Casual comments may appear in
print
8. Avoid bad or foul language.
9. Do not hide from the media. They can sense this and form an unfavorable
opinion of the press officer(s) as a credible source of news.
10. Do not answer questions beyond personal knowledge or expertise.
11. Do not permit media persons to attend emergency response team meetings.
These are likely to be technical meetings with lively discussions that may
last forever if people are performing rather than dealing with the problem at
hand.
Reasons for planning for and controlling statements made during a severe emergency
go beyond a simple desire to ensure orderly and accurate dissemination of information. One
of the groups showing up more frequently at hazardous materials accidents are lawyers
representing a wide variety of interests. It should be remembered that good lawyers will
remember everything they see and hear. A thoughtless comment or statement can surface
months later in a courtroom. Placements of blame, criticisms of response actions, airing of
dirty laundry in public, and similar statements can result in lengthy and messy legal battles
over comments or charges made in the heat of a very hectic moment.
14-53
-------
rrEM#sci
Topic: Plugging/Stopping of Leaks
When/Where Applicable: Any locale in which an industrial or transportation accident
may result in a leak or puncture of a tank, pipeline, or other
container.
Planning Goal: To ensure the jurisdiction of concern has at least a basic
capability of plugging or stopping leaks in tanks, pipelines, or
other containers.
Action Items:
• Assign responsibility to one or more individuals for identifying methods of
plugging or stopping leaks, assembling the materials and supplies necessary
for this task, and training for their use under emergency conditions.
• Alternatively, identify and arrange for the rapid provision of the above
services on an emergency basis by an expenenced and qualified pnvate
contractor.
Discussion:
Small leaks left unattended for extended penods of time can cause large losses of
chemicals to the environment and much more severe effects than would occur if the leak was
somehow completely or partially plugged on a prompt basis or some other means were
employed to reduce outflow of the hazardous material. There are great benefits, therefore, in
having access to one or more individuals with the basic tools and knowledge needed to limit
losses from punctured or leaking tanks or pipelines
The most widely available means for plugging holes or leaks in equipment involves use
of conical, cylindrical, square or wedge shaped pieces of wood, rubber or metal sheets,
inflatable pipe plugs, pneumatic leak sealing "bandages", special patching compounds,
clamps of various types, and a number of other items. The plugs alone, if available in a
variety of sizes, can be jammed into holes and greatly reduce the open area from which the
contents of the tank or pipeline can escape; assuming, of course, that it is safe for individuals
to approach the leak area The book by J R. Cashman listed at the beginning of this chapter
is especially enlightening on this topic. Several vendors market special leak plugging and
patching fats. Innovative response personnel may be able to fashion their own devices.
Where a tank vehicle is losing liquid cargo, it may be worthwhile to have the means
available to turn the body of the vehicle over such that the point of leakage rises This will
lessen the total amount of cargo than can escape before the liquid level in the tank drops
below the height of the hole Be advised, however, that this may require special equipment,
trained personnel, and expert supervision for a safe outcome.
14-54
-------
Various types of transportation containers have internal emergency shutoff valves that
can reduce or stop outflow from external valves that have been damaged m an accident.
Knowledge of the standard placement and use of these valves can be invaluable, most
particularly for highway tank vehicles.
Many incidents are brought to a rapid end simply by having the proper common tools
available to close a valve or tighten some bolts
14-55
-------
ITEM#SC2
Topic: Suppression of Hazardous Gas or Vapor Releases
When/Where Applicable: Any locale where large quantities of toxic or flammable vapors
may be released to the atmosphere.
Planning Goal: To ensure readiness to take rapid measures to reduce the rate
or amount of hazardous vapors or gases that enter the
atmosphere in an accident.
Action Items:
• Use hazard analysis results to identify all major potential sources of
hazardous gas or vapor releases into the atmosphere
• Identify the specific hazardous materials that may be discharged (to the
extent possible).
• Select appropriate vapor or gas hazard mitigation measures for each
significant threat
• Plan for rapid availability of qualified manpower, special equipment,
materials, or supplies necessary to mitigate gas or vapor hazards.
Discussion:
Gases or vapors may enter the atmosphere directly from broken, ruptured, or punctured
containers, or alternatively, from evaporating or boiling pools of liquid that have been
discharged to the environment. There are several response measures beyond plugging or
stopping the leak source (discussed elsewhere) that may be used to reduce the rate or amount
of airborne contamination. These include:
• Physical restriction of liquid pool surface areas,
• Use of firefighting or specialized hazardous material foams on liquid pools,
• Dilution or coverage of liquid pools with water (or other compatible and
safe liquids),
• Use of water sprays or fogs,
• Neutralization of spilled liquids,
• Cooling of spilled liquids or venting tanks, and
• Intentional ignition.
14-56
-------
Chapter 3 of this guide reported and explained the observation that the total rate of
vapor or gas evolution from a pool of liquid is directly related to the surface area of the pool
Thus, there are benefits to physical restriction of exposed pool surface areas. This can be
accomplished by rapid containment of discharging liquids by building dikes or other barriers
to flow or digging of trenches or sumps. The goal should be to keep the top surface area of
the pool as small as possible
Once the pool has been contained, other methods may be used to reduce the area
exposed to the atmosphere The most common response is to cover the surface of the liquid
with a compatible foam, since a thick foam blanket may in many cases greatly reduce gas or
vapor evolution, even from pools of quiescent liquefied gases. Note, however, that the
application of a warm foam to the surface of a cold pool of liquid may result in even greater
gas or vapor evolution for an initial penod of time
When the spilled liquid has a normal boiling point above ambient environmental
temperature, is completely or partially soluble in water, and is not dangerously reactive with
water, its vapor pressure and therefore its evaporation rate can be reduced by diluting the
liquid with large amounts of water Reductions in the evaporation rate will then result in a
smaller downwind hazard zone.
The same technique can also be used for soluble liquids with normal boiling points
below the ambient temperature and liquids that generate heat upon contact with water
However, since the introduction of water to such a pool may actually increase gas or vapor
generation while water is being introduced, this action should only be undertaken with great
caution It is best for use when the public has been cleared from vulnerable downwind areas
and there is a desire to reduce the time duration over which the pool would otherwise pose a
downwind hazard.
There are some liquid hazardous materials that are insoluble, heavier than, and
non-reactive with water Once contained, vapor evolution from these liquids can be reduced
or eliminated by carefully covering the liquid pool surface with a layer of water that will float
on the contained liquid Occasionally, the same principle can be applied using a compatible
and safe liquid other than water
Many spill response guides suggest the use of water sprays or fogs (from fire hoses and
nozzles) to knockdown, absorb, or disperse hazardous vapors in air If the spilled liquid is
dangerously reactive with water, plan to apply the fog or spray at a point sufficiently
downwind of the spill point so that water will not contact the pool If there are hazardous
liquid fumes or aerosols in sax, or the gas or vapor is soluble in water, give some
consideration to the need to contain the possibly contaminated water runoff for later
collection and disposal Finally, realize that the application of water fog or spray to
flammable vapors or gases will not necessarily eliminate their fire hazard
Several hazardous materials, as discussed in Chapter 7 of this guide, can be neutralized
via a chemical reaction to one or more substances that pose lesser threats to public health or
the environment Where this response may be appropriate for a particular hazardous
14-57
-------
material, the response plan should identify sources of necessary neutralization agents and the
means to apply them to spilled substances. Solid neutralization agents may possibly be
"applied" with snow blowers or seed spreading equipment. Liquids may be applied with
spraying equipment. In all cases, pay attention to the fact that the neutralization agent may
itself be hazardous hi some fashion and that the reaction may potentially produce heat or be
violent if improperly controlled. Seek expert advice and assistance where necessary
Since the evaporation or boiling rate of a liquid is a function of its vapor pressure, and
since the vapor pressure is a function of temperature, there may be situations in which the gas
or vapor hazard can be reduced by cooling spilled liquids or the containers from which gases
are venting This might be accomplished using large quantities of ice or dry ice Where
available and where the means exist for safe use, liquid nitrogen may also be considered, as
may supplies of carbon dioxide. Consult qualified experts for advice to determine if this is a
viable and safe option to consider for response purposes
The last measure to be discussed is not one that is often suggested in spill response
guides because its application can often cause additional complications and hazards in
densely populated or industrialized areas Nevertheless, where a highly toxic gas or vapor is
being released to the atmosphere, the resulting gas or vapor cloud or plume can cause (or is
causing) widespread deaths and/or senous injuries, and the gas or vapor is flammable or
capable of being ignited, give careful consideration to the possibility of intentional ignition of
escaping hazardous materials, possibly from a distance using a flare gun or other means of
ignition. This action can greatly reduce the toxic hazard of the escaping gas or vapor in
many cases. However, depending on the circumstances and surroundings of the discharge,
intentional ignition can also result in flame jets, tank BLEVEs, fireballs, vapor or gas cloud
fires or explosions, or pool fires that may themselves cause severe problems Thus, this must
be considered a "last ditch" response not to be undertaken without due consideration of its
implications and ramifications.
14-58
-------
ITEM#SC3
Topic' Containment of Spills of Liquids or Solids on Land
When/Where Applicable: Any locale where large quantities of liquid or solid hazardous
materials may spill on land in an area without secondary
containment systems
Planning Goal. To ensure that equipment, materials, and supplies are available
to contain spills of hazardous liquids or solids on land
Action Items.
• Where necessary, arrange for rapid availability of bulldozers or other
earthmoving equipment capable of building dikes or digging trenches
• Where necessary, arrange for rapid availability of properly equipped work
crews with shovels or other equipment to build dikes or dig trenches.
• Where necessary, arrange for rapid availability of plastic sheeting or other
compatible materials that can be used to cover spilled solids (to prevent
agitation by the wind or wetting) or to line dikes, basins, or trenches used to
collect liquids
• Where necessary, plan for rapid sealing of drains and sewer openings to
prevent entry of hazardous materials.
• Where necessary, plan for the rapid plugging of sections of storm drains to
limit the spreading of hazardous materials that have entered a drainage
system (Note. Plan for rapid access to a map showing the layout of local
systems. See the last paragraph of the discussion section.)
Discussion:
One of the first steps in spill response when a liquid or solid has been discharged onto a
land surface is to attempt to contain the spilled material and to prevent the further spread of
contamination Although specialized equipment has been developed to construct dikes of
foamed concrete or plastic materials, the most widely available and generally adequate
substances to use are earth, sand, clay, and plastic or rubber sheeting.
Dikes or barriers of earth, sand, or clay materials can be quickly constructed with
bulldozers, similar equipment, or properly equipped individuals with shovels Note, howev-
er, that motonzed equipment should not be used indiscriminately in the vicinity of flammable
or explosive vapors or gases Note also that response personnel may require special
protective clothing and breathing apparatus to approach a spill.
14-59
-------
Plastic or rubber sheeting can be used to cover spilled solids This can prevent the
wind from causing toxic dusts to become airborne and also protect the bulk of the solid from
becoming wet from rain or hose streams used in the area, thus reducing the extent or
possibility of soil, groundwater, or surface water contamination by the spilled substance.
Such sheeting can also be useful for lining dikes, basins, and trenches for similar purposes
where liquids are to be contained. Finally, sheeting materials, together with stones or bricks
and soil, sand or clay, can be used to cover storm drain openings in a pinch
When hazardous materials have already entered a storm drain system, there are benefits
to attempting to limit or contain the flow of the material by damming at strategic locations
Since storm drainage systems typically flow into bodies of surface water, containment might
prevent significant water pollution and facilitate later cleanup However, note that at least
one authority suggests that volatile chemicals should never be trapped in a closed conduit
such as a storm drain, probably due to the possibility that explosive vapors may accumulate
and encounter a source of ignition. Thus, judgments on damming must be made on a
case-by-case basis, with special attention being given to cases in which contaminants may
enter a water treatment plant
14-60
-------
ITEM#SC4
Topic: Cleanup of Spills of Liquids or Solids on Land
When/Where Applicable: Any locale where large quantities of liquid or solid hazardous
materials may spill on land.
Planning Goal: To ensure that equipment, materials, and supplies are available
to recover spills of hazardous liquids or solids on land.
Action Items:
• Where necessary, arrange for rapid availability of pumps, hoses, and
temporary storage containers (or alternatively, vacuum trucks) to recover
pools or other accumulations of hazardous liquids.
• Where necessary, arrange for rapid availability of properly equipped work
crews with appropriate equipment.
• Where necessary, arrange for rapid availability of drums or other containers
to hold contaminated solids, soil, or leaking packages.
• Where necessary, arrange for rapid availability of absorbent materials.
• Where necessary, arrange for rapid availability of front-end loaders and
other earthmovmg equipment, including dump trucks.
Discussion:
Once a spilled substance has been contained, the next step is to remove it from the
environment. The effort is often undertaken by the parties responsible for the spill or
discharge or a spill cleanup contractor it may hire, but local, county, and state governments
should have a capability to respond when the responsible party is unknown or is unprepared
or unwilling to take action and circumstances do not permit waiting for federal intervention
under CERCLA, SARA, or the Clean Water Act. The planning effort should involve. 1) a
review of federal contingency plans prepared by regional EPA and or U.S. Coast Guard
Offices as well as the state agency primarily concerned with environmental protection to
learn how these agencies plan to respond when needed; 2) a decision as to whether spills will
be cleaned up by local government personnel or by specialized cleanup contractors hired by
public authorities; and 3) arrangements for rapid availability of necessary services, equip-
ment, and supplies.
The methods usually applied for "gross" cleanup of contaminated ground surfaces are
rather straightforward They involve
1) Use of compatible pumps, hoses, and tanks, drums, or vacuum trucks to
collect pools of accumulated liquids;
14-61
-------
2) Use of soil, sand, clay, straw, sawdust, fly ash, cement powder, perlite,
vermicukte, or commercially available mineral or plastic sorbent materials
to absorb and mop up liquid residues, and
3) Removal of contaminated surface layers (where cleaning in place is not
practical) by shovels or mechanical means
In all cases, of course, supplies and equipment used must be compatible with the
hazardous material spilled, workers must be protected from toxic exposures, and special care
must be taken in the presence of potentially flammable or explosive atmospheres.
Other current cleanup techniques include steam-cleaning or detergent washing of solid
surfaces where hydrocarbons or similar materials have spilled, burning of flammable or
combustible materials in place where it is safe to do so and permitted by regulatory
authorities, and application of special mutant bacterial cultures to contaminated soils and
liquids to "digest" contaminants over a penod of time.
14-62
-------
ITEM#SC5
Topic- Containment of Spills into Water Bodies
When/Where Applicable: Any locale where significant quantities of hazardous materials
may spill into a body of surface water
Planning Goal: To ensure that equipment, materials, and supplies are available
to contain spills of hazardous liquids or solids into bodies of
water.
Action Items-
• Use hazard analysis results to determine the specific materials that may be
spilled into a body of water (to the extent possible).
• Where necessary, arrange for rapid availability of waterborne spill contain-
ment equipment and supplies.
• Where necessary, arrange for rapid availability of earth moving equipment,
boats, spill containment booms, sorbent materials, sand bags, and other
potentially necessary items
Discussion
The selection of a water-spill containment method for any given substance mostly
depends upon how the substance will behave when spilled into water The potential behavior
of such spills was discussed in Chapter 3 of this guide
Spills of lighter-than-water and mostly insoluble materials are best handled with spill
containment equipment and methods These include use of commercial or home-made oil
spill booms, chemical spill herders, or use of hose streams or small-boat propeller washes to
control the spread of the substance on water In addition, where small rivers or streams are
affected, underflow dams may be constructed, these being dams with open pipes well below
the surface of the water such that water passes through the dam but floating materials are
trapped Oil spill response guides will address these and other potential containment
methods in more detail Due to the specialized nature of the equipment and materials
necessary, most jurisdictions are probably best advised to find a local oil spill containment
and cleanup contractor willing to respond on an emergency basis for these types of spills
Spills of heavier-than-water and mostly insoluble substances will sink to the bottom of
a water body Although it is not always effective or possible, consider taking advantage of
natural deep-water pockets, using sand bags, or building low underwater dams or dikes to
trap the liquid on the bottom and prevent further downstream contamination
14-63
-------
Substances that are partially or completely soluble in water are the most difficult to
contain. At best, if the body of water is of a manageable size, an attempt can be made to dike
the upper and lower bounds of the body of water and to divert incoming flows of clean water
around the contaminated area using earthmovmg equipment This is easier said than done m
many cases but has been successfully attempted at times.
14-64
-------
ITEM#SC6
Topic- Cleanup of Spills into Water Bodies
When/Where Applicable: Any locale where significant quantities of hazardous materials
s may spill into a body of surface water.
Planning Goal' To ensure that equipment, materials, and supplies are available
to remove or mitigate contaminants in bodies of water.
Action Items
• Where necessary, arrange for rapid availability of oil spill cleanup
equipment and services
• As and if necessary, arrange for rapid availability of chemical spill
treatment and cleanup services.
Discussion.
A variety of methods and equipment have evolved over the years for treating or
removing spills of oil-like substances in water. These include use of oil skimming
equipment, sorbents, burning agents, sinking agents, and dispersants. The planning effort for
such materials should involve a review of federal contingency plans prepared by regional
EPA and/or U S. Coast Guard offices, as well as the state agency primarily concerned with
environmental protection, to learn how these agencies plan to respond to major oil spills
Given the special equipment, materials, and training needed for the effort, it is probably best
recommended that arrangements then be made with a local oil spill cleanup contractor for
emergency response (Note' There are also advantages to be realized by coordinating and
integrating this effort with state and federal plans Indeed, many jurisdictions may find that
state and/or federal plans fully cover them for this type of spill and that no specialized
planning is necessary. Others may discover local oil spill cooperatives or other industry
groups willing to lend assistance even if a member company is not responsible for the
incident)
The situation gets a bit more complicated where spills of water-soluble materials or
heavier-than-water insoluble materials are possible, since the equipment and methods
necessary to remove or treat contaminants in the water are even more specialized and
sophisticated. Again, it is best to start the process with a review of state and federal
contingency plans and to integrate and coordinate planning efforts for these types of
materials Note, however, that any potential contractors identified should have experience
with carbon adsorption, chemical neutralization, and other water treatment technologies.
These highly specialized contractors are not as easy to find as oil spill cleanup companies
that occasionally handle chemical spills, and indeed, may not be found at all within the
borders of some states
14-65
-------
ITEM#SC7
Topic: Support Services for Field Response Forces
When/Where Applicable: Any locale that may experience a hazardous material spill or
discharge requiring a prolonged response action in the field
Planning Goal: To provide food, rest areas, and other equipment and supplies
necessary to sustain field response forces.
Action Items:
• Where necessary, arrange for canteen services, accommodations, electrical
power, lighting, heating, portable toilets, washing facilities, and other
supplies and services needed to sustain workers at a spill site
Discussion:
Several situations can be envisioned in which response personnel m the field may be
required or needed at the site of an accident for more than a day These personnel will
require rest areas and food to be able to continue to function, particularly if they are in a
remote area or a region that has otherwise been evacuated Sources of assistance for
planning in this area include the American Red Cross, the Salvation Army, and the National
Guard or Army Reserve. The latter organizations in particular might be able to provide tents,
cots, fans or heaters, lights, and expertise in layout and management of staging or rest areas
Although the above individuals and their command personnel must be given priority, it
may also be prudent to consider how to assist media personnel and government officials from
outside the jurisdiction in finding appropriate lodging At one protracted incident in Florida
some years ago, 223 representatives of different federal, state and local agencies were present
™ all of them watching a crew of 21 people cleaning up the scene. Add to these media,
industry, and other personnel involved in the overall emergency response, and it is easy to
see that the total number of "strangers" that may be on scene can be substantial
The OSHA regulations involving a site safety and health plan (see Item #CC4) contain
numerous requirements pertaining to support of field personnel and should be referred to for
further guidance. Be especially aware of the need to ensure that established hygiene
practices preclude the possibility of mgestion of contaminated food and refreshments
14-66
-------
ITEM#SC8
Topic: Maintenance of Apparatus and Equipment
When/Where Applicable: Any locale that may experience a hazardous material spill or
discharge requiring a prolonged response action in the field.
Planning Goal: To provide fuel, engine fluids, and maintenance services to
support the on-scene operation of apparatus and equipment
operating in hazardous environments.
Action Items:
• Use of the results of the hazard analysis to identify situations where
apparatus and equipment operating in contaminated areas may require
refueling and minor maintenance to keep them operating
• Establish arrangements for obtaining fuel, engine fluids, and other mainte-
nance items (e.g, belts, filters, etc.) to support continuous operation of vital
apparatus and equipment
• Establish procedures for providing refueling and maintenance services to
on-scene apparatus and equipment operating in potentially contaminated
areas
Discussion:
During long-duration incidents, apparatus (e.g., fire department units, backhoes, etc)
and equipment (e g, generators, pumps, etc.) may require on-scene refueling and minor
maintenance to enable uninterrupted operation On-scene services of this nature are
beneficial in that they eliminate the need to remove deployed apparatus and equipment from
the incident scene By refueling in-place, operations can continue uninterrupted and the need
to replace chemically exposed equipment with uncontaminated items is eliminated, thus,
reducing the number of units potentially requiring decontamination. Note that operations
must be conducted safely and in such a manner that will prevent the spread of contaminants
from response vehicles and equipment to maintenance vehicles and refueling tanks
Furthermore, it may be necessary for maintenance personnel to wear appropnate body and
respiratory protective gear when operating in hazardous environments Also note that
provision of refueling and maintenance services may not be prudent in particularly hazardous
locations For example, fire department apparatus positioned near a fire should not be
refueled where ignition of fuel vapors is possible One step that may possibly eliminate the
need to refuel apparatus and equipment is to fill fuel tanks to capacity pnor to deployment.
14-67
-------
ITEMtfSDl
Topic: Documentation of Response Activities and Costs
When/Where Applicable: Any locale that may find it necessary to undertake a major
response action due to a spill or discharge of hazardous
material.
Planning Goal: To ensure that a careful record is maintained of what happened
and what was done in response.
Action Items:
• Assign responsibility for realtime and post-incident documentation of the
accident and resulting response actions
• Create or adopt appropriate reporting forms and procedures.
• Arrange to collect the records from various sources in a central and safe
location.
Discussion*
Detailed records of what happened and what actions were taken m response can help
in:
• Attempting to recover response costs and damages from the party responsi-
ble for the incident
• Setting the record straight where there are charges of negligence or
mismanagement resulting from the incident
• Reviewing the efficiency and effectiveness of response actions.
• Preparing for future incident responses
• Verifying facts, actions, injuries, equipment used, etc for the purpose of
legal proceedings, insurance claims, budget requests, and public inquiries
In addition to written documentation of an incident, it is good practice to draw
diagrams or sketches of containers, vehicles, structures, streets, containment techniques
employed, locations of deployment, etc. Photographs and videotapes should also be kept on
file for reference purposes.
14-68
-------
ITEMtfSMl
Topic: Monitoring of Atmospheric Conditions
When/Where Applicable: Any locale with a reasonable potential to experience a major
release of hazardous vapors or gases into the environment
Planning Goal: To ensure that command personnel have continuous access to
current data, predictions, and expert advice regarding weather
and atmospheric conditions.
Action Items.
• Identify sources of weather data and predictions in the immediate area
• Arrange for continuous updates of weather information during incidents
involving major discharges of hazardous vapors or gases into the environ-
ment
Discussion.
As discussions in Chapter 3 and 12 of this guide have shown, atmospheric stability
conditions, wind velocities, and wind directions have a direct impact on the boundaries of
downwind areas threatened by a plume or cloud of hazardous vapor or gas. Changes in these
conditions over time, particularly in the case of prolonged discharges, can require changes in
the boundaries of hazard zones Consequently, tracking of cloud or plume movements can be
greatly facilitated by direct access to a weather station manned by trained meteorologists.
Likely locations include major radio and television stations, major airports, and offices of the
National Weather Service, a part of the U.S. Department of Commerce.
Other atmospheric conditions of possible interest involve temperature, precipitation,
and humidity forecasts Temperatures can affect the physical state, vapor pressure, and other
properties of hazardous materials, as explained in Chapter 2, and also impacts on the length
of time a person can safely function inside a fully encapsulating protective suit. Precipitation
can impact dispersion of airborne contaminants, lead to runoff of contaminants in water,
cause dilution of spilled chemicals, and/or assist fire control efforts. Moisture in the
atmosphere may cause either adverse or beneficial chemical reactions involving spilled or
discharged materials.
14-69
-------
ITEM#SM2
Topic: Monitoring of Contaminant Concentrations
When/Where Applicable: Any locale where the air, water, or land may be contaminated
by toxic materials.
Planning Goal: To ensure rapid availability of the personnel and equipment
needed to sample, analyze, or otherwise monitor pollutant
concentrations in air, water, or soil.
Action Items:
• Contact the nearest EPA regional office and the state agency with primary
responsibility for environmental protection to assess their capabilities for
determining and monitoring pollutants in the environment during a signifi-
cant spill or discharge incident
• If these agencies cannot respond promptly in all cases or do not have a full
range of capabilities, particularly for air sampling and analysis under
emergency conditions, identify and arrange for the services of university
laboratories or commercial firms that have the necessary resources and
personnel and who are willing and prepared to respond
Discussion:
Determination of the concentrations of airborne contaminants at various points
downwind of a spill site can greatly help in determining the actual boundaries of hazard
zones and in deciding when reentry of these zones is feasible and safe. Similarly,
measurements of water or soil contamination can help determine the exact level of
contamination of these resources.
In order to assess the potential for adverse health impacts (which may not become
immediately apparent), it is often wise to plan for monitoring of contaminant exposures
experienced by response personnel and the general public as best possible under emergency
conditions.
Where potential exists for contamination of food and/or water supplies, responsibility
must be assigned for detection of such contamination via use of the resources identified
during this planning activity and for prevention of the consumption of known or potentially
contaminated food or water by people or animals.
14-70
-------
ITEM#SR1
Topic: Provision of Alternate Water Supplies
When/Where Applicable. Any locale with a water supply vulnerable to contamination by
hazardous materials
Planning Goal: To ensure rapid availability of potable water for affected
populations.
Action Items:
• Identify local sources of potable water supplies
• Determine if there is potential for contamination of these supplies in the
event of a hazardous material accident
• If there is potential, identify alternate sources of clean water and plan for its
distribution to residents
Discussion
There are a number of circumstances under which a potable water supply may become
unfit for human consumption for a time and require replacement This is most commonly
accomplished by bringing in supplies of bottled water and/or tank trailers capable of carrying
water The latter may be available from local National Guard units. Tank trucks that carry
milk are another possibility for consideration once thoroughly cleaned.
14-71
-------
ITEM#SR2
Topic: Cleanup of Dead or Contaminated Livestock or Wildlife
When/Where Applicable: Any locale where the discharge of a toxic gas or vapor into air
or the discharge of a toxic substance into water may result in
mass casualties among animal populations.
Planning Goal: To make provisions for the collection and disposition of the
bodies of dead aquatic and terrestrial animals and decontami-
nation and care of contaminated animals
Action Items:
• Identify a source of work crews and equipment for the collection and
disposal of dead annuals.
• Identify a disposal site where the bodies can be buried and or incinerated
• Identify a source of work crews and equipment for the decontamination and
care of contaminated animals.
Discussion:
A large number of animal carcasses, be they hundreds of cows or chickens or many
thousands of fish, can pose a variety of health hazards. Emergency response plans do not
typically address this topic, possibly because it is somewhat distasteful, but the topic is
worthy of consideration during the planning process where applicable. Where oil spills may
contaminate waterfowl, consideration might be given to the formation of work crews to
collect, clean, and care for the animals under the direction of experienced personnel. State
and federal agencies can provide assistance in planning for this latter activity Local
veterinarians and animal conservation groups may also be helpful, but in all cases, ensure
that personnel will not be placed at risk of adverse safety or health impacts by their actions
14-72
-------
ITEM#SR3
Topic: Post-Incident Testing for Contamination
When/Where Applicable: Any locale where toxic and potentially persistent contaminants
in air or in floodwaters may contaminate exposed surfaces
Planning Goal: To prepare for the potential need to check crops, ground
surfaces, homes, stored foods, and animals that may become
part of the human food chain for possible chemical contamina-
tion.
Action Items:
« Contact the nearest EPA regional office, the state agency with primary
responsibility for environmental protection, the regional Animal & Plant
Health Inspection Service office and/or the Food and Safety & Quality
Service Office of the U S Department of Agriculture, and other appropriate
agencies to assess their combined capabilities for contamination testing in
the aftermath of a significant spill or discharge incident.
• Where additional services may become necessary, identify and arrange for
the services of university laboratories or commercial firms that have the
necessary resources and personnel and who are willing and prepared to
respond.
Discussion.
To some extent, this item overlaps with Item #SM2, Monitoring of Contaminant
Concentrations. However, that item covers sampling and analysis activities during the initial
phases of a response action while this one is concerned with the potential for persistent
aftereffects.
14-73
-------
Topic: Structural Inspections after Fires or Explosions
When/Where Applicable: Any locale where a fire or explosion involving hazardous
materials may damage a large number of buildings, a bridge, a
tunnel, or other structures
Planning Goal: To make provisions for inspecting the structural integrity of
damaged buildings, bridges, tunnels, or other structures in the
aftermath of a fire or explosion.
Action Items:
• Identify the local, county, and/or state personnel with responsibility for
various types of structural inspections.
• Establish procedures, where a potential need is envisioned, for inspecting
damaged structures that may be contaminated with chemical residues.
Discussion:
A major explosion could damage or destroy numerous buildings and any nearby
bridges or tunnels. Similarly, large fires can have major effects over a wide area. In either
case, residents of partially damaged buildings will want to know if the structures are safe to
occupy while they await repairs Questions pertaining to the safety of highway or railway
bridges or tunnels must be resolved quickly to avoid traffic complications.
Note that inspection personnel may require special precautions (i e., chemical protec-
tive gear) in addition to normal safety equipment in those cases in which the structure may
still be contaminated by hazardous residues.
14-74
-------
ITEM#SR5
Topic: Provision of Post-Incident Recovery Services
When/Where Applicable: Any locale with the potential to experience a large number of
deaths or injuries and/or a large loss of residential properties
Planning Goal: To provide a central location where people can visit to seek
advice or help from social service organizations, government
agency representatives, legal aid sources, representatives of the
party responsible for the incident, and so forth
Action Items:
• Identify the key agencies and groups expected to have a role in post-disas-
ter recovery.
• Identify a convenient location where various organizations can set up shop
temporarily in the aftermath of a disaster.
• Plan to set up an information clearinghouse for answering questions from
the public.
Discussion:
There are numerous organizations that have a role in post-disaster recovery operations
Simultaneously, even if general guidelines are reported by the press, individual members of
the public will have specific questions on how to handle recovery of losses, repair of homes
and businesses, rerouting of mail, contact with "missing" relatives or friends, etc. These
action items set the stage for orderly provision of necessary services.
14-75
-------
ITEMtfTRl
Topic: Training of Response Personnel
When/Where Applicable: Any locale with the potential to experience significant
hazardous material accidents
Planning Goal: To ensure that key personnel have the necessary training to
properly conduct then: missions
Action Items:
• Based upon the types of incidents most likely to occur and the related
response and planning activities suggested by this chapter, determine the
types of training required for emergency response personnel with responsi-
bilities in any or all phases of the response.
• Determine what training the personnel have already had
• Identify and provide any new or additional training that might be required
• Conduct periodic drills to test the overall efficiency and effectiveness of the
emergency response plan and emergency response capabilities
Discussion:
There are a great number of duties and responsibilities associated with spill response
that differ significantly from the routine activities of public officials, police and fire
department personnel, and everybody from bulldozer operators to bus and ambulance drivers
who may have to function in difficult and possibly hazardous environments Appropriate
training can range from formal courses at pnvate spill control schools, universities, or
community colleges to training sessions provided by the local or state government in which
personnel are bnefed on their specific duties in an emergency and shown how to wear and
properly use personal protective clothing and devices The nature and extent of training
obviously depend on the nature of hazards to be faced, the specific responsibilities of
individuals, local and state training requirements, and federal training requirements specified
by OSHA (see 29 CFR 1910.120). Information on federal training opportunities is provided
in FEMA 134, the Digest of Federal Training in Hazardous Materials See Chapter 1 for
the address to write for obtaining FEMA publications
14-76
-------
ITEM#WD1
Topic: Disposal of Hazardous Wastes
When/Where Applicable: Any locale where recovered chemicals and/or contaminated
materials may require proper disposal
Planning Goal' To identify appropriate disposal sites for waste chemicals and
contaminated materials.
Action Items:
• Identify state and federal regulations pertaining to transportation and
disposal of hazardous wastes
• Identify authorized waste disposal sites.
Discussion*
Where the parties responsible for a spill do not take appropriate action, it may become
necessary for public authorities to undertake disposal of hazardous wastes. This requires
knowledge of waste disposal regulations, the location of approved and authorized disposal
sites, and the proper procedures for transporting and transferring wastes to these sites. Local
governments should seek assistance from the state environmental protection agency, the
EPA, and the Coast Guard for assistance with these efforts. These agencies will have
considered these problems in their own statewide or regional emergency response plans.
14-77
-------
-------
APPENDIX A
A TUTORIAL ON FUNDAMENTAL MATHEMATICAL SKILLS
Purpose of Appendix A
Various sections of this guide, literature sources that may be accessed to obtain the
information and data necessary for the conduct of a hazard analysis, and certain input
parameters required by the computer program provided with this document may require the
user to perform various computations, convert the units of certain numerical values, and read
various graphs. The purpose of this appendix is to provide a bnef tutorial on:
• Basic algebra
• Conversion of units
• Scientific notation
• Reading of log-log and semi-log graphs
• Surface area and container volume estimation methods, and
• Methods to estimate container content weight.
Each of these topics is discussed below for those who may not be familiar with them or may
have become a bit rusty on their use since their school days
A Review of Basic Algebra
Algebraic equations can then be thought of as sentences of instructions, where simple rules
replace the words The main symbols you may come in contact with, and their meanings are:
A/B or A means "A divided by B"
B
A x B or (A)(B) means "A times B"
A + B and A - B respectively mean "A added to B" and "B subtracted from A"
( ) means treat everything inside these parentheses as one variable, or equivalently,
perform all computations inside parentheses before performing other computations.
A = B means "A equals B"
A < B means "A is less than B"
A > B means "A is greater than B"
A-l
-------
A £ B and A < B respectively mean " A is greater than or equal to B" and "A is less
than or equal to B".
A1 means A raised to the B power
If parentheses don't guide you, mathematical operations are performed in the following
order, left to right
1) Raise all powers first
2) Perform all multiplications or divisions second
3) Perform all additions or subtractions third
Some examples of the types of equations that may be encountered in this guide and
elsewhere include:
A=2x60 = 120
B =A x 10x4= 120x 10x4 = 4800
r 10-6 4
C~4-2~2~2
= (1.8x2)+32 =
A Review of Unit Conversions
It is important that the numbers used hi equations or input to a computer program are
provided in the proper sets of units For instance, if the spill amount to be entered into an
equation or computer program is desired in pounds, the program will produce wrong answers
if the user enters a number expressing this amount in tons. This may seem obvious, but it is a
frequently made mistake at all levels of experience. Checking that units are correct (called
dimensional analysis) is a straightforward procedure, but still one that requires a little
thought.
All mathematical methods and procedures presented in this guide and its associated computer
program specify the units in which any values (such as length, weight, or area) must be
provided. When the user has these values in the proper set of units, then he or she can use
them immediately. However, if one or more of the values are in the wrong units, the values
must first be "converted". Although the computer program provided with this document
contains a units conversion utility to assist the user in these tasks, it is still a good idea to
know how unit conversions are generally accomplished
A-2
-------
Conversions usually involve the multiplication or division of a value in one set of units by a
"conversion factor" to change it to another set of units. Some simple and mostly self-ex-
planatory examples of the process include:
IQOyards x - = 300/eef
Imiles x528Qfm==10,560/egf
mile
IQOgallons x— — = 13.37ft3
7 48gallons
13.37#3x-
^xn ii 14.7 psia ^nn
640mmHg x— - —x— ^ — =12 3&psia
760nvnHg latm
ton
Table A.1 lists these and other simple conversion factors that are likely to be most often
needed. Note that in each of the above examples the "net" units or both sides of the equal
sign are (and should be) the same after units appearing both on the top and bottom of an
expression are cancelled out Special equations for temperature conversions can be found in
Section 22 of this guide Conversions between "absolute" and "gauge" pressure are
discussed in Section 2 3.
A Review of Scientific Notation
Numerical shorthand called "scientific notation" makes it easier to work with numbers that
are very small or very large Basically, it involves writing a number down (like 3 6) and then
showing how many times this number must be multiplied or divided by 10 to obtain the
number that is of interest to the reader For example, assume that the number 360 is to be
written in scientific notation This number would be represented by 3 6 x 102, where the 2
above the 10 means that 3 6 must be multiplied 2 times by 10 (3 6 x 10 x 10 = 360)
For small numbers less than 1.0 (like 0 036), a minus sign is placed hi front of the number
above the 10 (this number being the "power" of 10) to show how many times the first
number must be divided by 10. Thus, 0 036 would be shown as 3 6 x 10-* (3 6/10 = 0 36,
A-3
-------
TABLE A.1
SELECTED UNIT EQUIVALENCY FACTORS
Length
1 mile = 0.62 kilometers
1 mile = 5280 feet
1 nautical mile = 6080 feet
1 kilometer = 1000 meters
1 yard = 3 feet
1 meter = 3.281 feet
1 foot = 12 inches
1 inch = 254 centimeters
Pressure
1 atmosphere = 14.7 psia
1 atmosphere = 760 mm Hg
1 atmosphere = 101,325 Pascals
14.7 psia = 760 mm Hg
Weight*
1 ton = 2000 pounds
1 kilogram = 2.2 pounds
1 kilogram = 1000 grams
1 pound = 454 grams
1 gram = 1000 milligrams
Energy
1 Btu = 252 calories
1 lolocalone = 1000 calories
1 Btu =1055 joules
Volume
1ft3 = 7.48 gallons
1m3 = 35.32 ft3
lyd3 = 27.0 ft3
*Assuming gravitational acceleration at sea level
A-4
-------
0.36/10 = 0.036). Every time a number is multiphed by 10, its decimal point moves one
position to the right When it is divided by 10, the decimal point moves one position to the
left
The following table should help in gaining a better understanding of the concept
Number
0.000036
0.0078
0.510
9.2
51.0
780.0
36,000.0
Scientific Notation
3.60 x 10s
7 80x10*
5.10 x 10 l
9.2 x!0° (10° =1)
5.1 x 101
7.8 x 102
3.6 x 10*
Once this basic concept is mastered, it is necessary to understand how to add, multiply or
divide two numbers written in scientific notation This gets a bit more difficult. In all cases,
it is important to keep track of the correct powers of 10.
The easiest way to add two numbers in scientific notation is to convert them back into
numbers in standard notation. For example, two numbers like 3.6 x 102 and 2.2 x 103 become
360 and 2,200. Adding them together gives 2,560 or 2.56 x 103. The numbers 7.8 x 103 and
5.1 x 103 become 7,800 and 5,100. Then- sum is 12,900 or 1.29 x 10*. Likewise, 9.2 x 104
and 5.1 x 10* become 0.00092 and 0.051, and the sum is 0 05192 or 5.192 x 10-2.
The same technique can be used for multiplication, but there is another way that some might
find easier. It involves multiplication of the first numbers in the scientific notation, and then
addition of the powers of 10 involved. For example:
(3.6 x 102) x (2.2 x 103) = (3.6 x 2.2) x 10»*3>
= 7.92x10*
(7.8 x 103) x (5.1 x 103) = (7.8 x 5.1) x 103*3)
= 39.78 x 10s
= 3.978 x 10>
A-5
-------
(9.2 x 1(H) x (5.1 x = (9.2 x 5.1) x 10* *>
10*) = 46.92x10*
= 4.692 xlOs
In division, the front numbers are divided and then the power of 10 for the bottom number is
subtracted from the power of 10 for the top number.
2.2 xlO3
= 1.64xlO~1
5.1 xlO3 5.1 xlO3
= 1.53x10°
= 1.53
9.2X10- 92X10-
5.1 xlO'2 S.lxKT2
x
= 180xlOH+2)
= 1.80 xlO"2
In the last example, it is important to note that two minus signs equal one plus sign.
Subtracting -2 from -4 is the same as adding 2 to -4 to get an answer of -2.
Finally, it is necessary to note that computer programs use what is referred to as "E format"
to represent numbers in scientific notation. In this format, the letter "E" is used to replace the
term "xlO" and the powers of 10 simply follow the E without appearing as superscripts.
Thus:
1.23 x 10« becomes 1.23E 06
8.7654 x 10J becomes 8.7654E-05
A-6
-------
How to Read Log-Log Graphs
Almost everyone at some time or another has been shown how to use a graph with normal
axes (the axes are the lines on the bottom and left side of the graph). This type of graph
usually resembles something like:
Curve
12
B
Given a known value for A on the left vertical scale or axis, the user would move to the right
horizontally until he or she meets the curve, and then would go straight down vertically to
find the appropriate value for B on the bottom scale of the graph. This procedure is
demonstrated by the dotted lines and arrow heads in the above drawing. If the value for B
was known instead of A, the value of A could be determined by following the dotted lines in
the reverse direction The reason this type of graph is said to have normal axes is that the
gradations on the side and bottom are equidistant The point for 2 on the B axis is the same
distance from 4 as 8 is from 10.
Another type of graph that is used when the ranges on the axes encompass many
powers of 10 (from 0.10 to 1,000, for example) is called a log-log graph. In this sort of
graph, the axes have logarithmic scales, and do not have gradations which are equidistant
They are a bit more difficult to read than graphs with normal axes, but permit much more
information to be placed on a single page.
A-7
-------
Figure A. 1 is an example of a log-log plot similar to many which may be encountered.
To understand how the graph is used, it is necessary to first review:
1. The gradations on the bottom B axis;
2. The gradations on the side A axis; and
3. The curves with different values for C.
The bottom B scale or axis spans 2 powers of ten; from 0.1 to 1.0, and again from 1.0
to 10.0. The arrows show where the intermediate numbers are between 0.1 and 1.0 to
demonstrate what a logarithmic scale looks like Note that the distance for 0 1 to 0 2 is much
greater than the distance between 0.8 and 1.0, but that someone can still tell what each
gradation means. The second half of the scale (to the right of 1.0) is broken up in the same
way.
The left vertical A axis also spans 2 powers of ten, but has markings in scientific
notation for numbers which are smaller than 1.0. The biggest number shown is 1 x 10-2, at
the top of the scale, and the smallest is 1 x Ifr* at the bottom. Again, the arrows show the
relative locations of intermediate points. It is important to note that two sets of numbers are
given on the graph for points at the very top, exactly in the middle, and at the very bottom.
That's because 10 x 10-3 equals 1 x Ifr2,10 x 10-* equals 1 x 10-3, and so on. Graphs you may
encounter elsewhere may not have such dual markings, so this fact should be remembered.
To add another "dimension" to this particular graph, three curves are shown for
different values of C. Many graphs in the literature will have more than one curve or line on
them so that more information can be contained in each.
To use the graph, one has to know values for two out of the three parameters that have
the relationship shown on the graph. For example, if it is known that A has a value of 2 x
10-3 and that C has a value of 4, the small arrows on Figure A 1 show how one would read
across and down to find the previously unknown B value of 0.35.
One problem with using a graph with many curves for different C values is that the
known value for C (or whatever else it is called) often lies in between two of the curves, or
above or below one of those on the outer boundaries In such a case, the user must estimate
where the correct curve would be if it were shown.
How to Read Semi-Log Graphs
Many graphs may be semi-log graphs. This means that the vertical axis on the left has a
logarithmic scale while the horizontal axis on the bottom has a normal axis Such graphs are
just a combination of the two types previously described
A-8
-------
10 X 10"3/1 X 10"2
8X10"3
6 X 10"3
4X10"'
3 X 10"'
2X10
,-3
10X10"4/! X10"3
10 X 10"5/1 X W4
^
*
ITS^
\
X
^v
>^_
>
0.1
•*-
N
N.
>y
>
j
02
V
X
s^
r
V
\
C = 5
\
03
\
s
\
\
1
04
0
\
\
^
s
6
^
s
s
s
,,
S
i
oe
C = 3
\
\
y C = 4
\
V
\
\
<
S
s
s
\
S
s
\
\
V
\
s
s
\
\
'
B
1C
1 0
FIGURE A.1
A TYPICAL LOG-LOG GRAPH
A-9
-------
Surface Area and Container Volume Estimation Methods
It may be necessary at times for the reader to determine or estimate the surface area of the
resulting pool when a liquid is discharged upon the ground. It is well, therefore, to review
how the area of a surface can be calculated from its dimensions. Of interest are circles,
squares, rectangles, and triangles. The equations used to compute the areas of these shapes
are:
For circles: Area = 3.1416 x radius x radius, or
Area = 0.7854 x diameter x diameter
For squares: Area = length of side x length of side
For rectangles: Area = length x width
For triangles: Area = 0.5 x length of base x vertical distance from base to
opposite tip of the triangle
There are several instances where the computer program associated with this guide ask the
user for the diameter of the hole in a tank or pipeline or of a liquid pool, even though it is
clear that pools or holes are not always circular. In such cases, the user of the methods is
expected to determine the diameter of a circle that has the same area as the area of the hole or
liquid pool. This is accomplished by using the equations:
Radius = /Area.
V 3.1416
Diameter^ 2 x Radius
There may be situations in which it is desired to determine the volume of a pipeline, tank, or
other container. Equations for volume estimation are:
For spheres: Volume = 4.19 x radius x radius x radius
For vertical cylinders: Volume = 3.1416 x height x radius x radius
For pipelines and Volume = 3.1416 x length x radius x radius
horizontal
cylinders:
Methods to Estimate Container Content Weight
A-10
-------
Once the volume of a container is known, it may be necessary at times to determine its
capacity to hold a liquid in units of pounds. It is mostly commonly expected that the user
will know the volume in units of gallons or cubic feet and will have access to the liquid
specific gravity or liquid density of the contents. Equations to determine weight in pounds
from different combinations of these values include:
Weight (Ib) = volume (ft3) x hquid density (lb/ft3)
Weight (Ib) = volume (ft3) x (62.4 lb/ft3) x liquid specific gravity
Weight (Ib) = volume (gallons) x (1 ft3/7.48 gallons) x liquid density (lb/ft3)
Weight (Ib) = volume (gallons) x (62.4 lb/ft3) x (1 ft3/7.48 gallons) x specific
gravity
A-ll
-------
-------
APPENDIX B 1
TECHNICAL BASIS FOR CONSEQUENCE
ANALYSIS PROCEDURES
B.I Introduction
This appendix provides a technical overview of the various computational
procedures contained in ARCHIE and their key assumptions Many of
these models reflect the state-of-the-art in terms of development work under-
taken at public and private research facilities, information in the published
literature, and data collected or estimated from experiments and actual ac-
cidents Nevertheless, to make models workable for ARCHIE within con-
straints imposed by available data and project resources, it was necessary
at times to employ a number of idealizations and assumptions which must
not be overlooked in the interpretation of results Accordingly, it is believed
that trends are predicted accurately, but estimated values are less certain
In general, estimates of hazard zones are conservative, in that uncertainties
overpredict rather than underpredict the extent of the hazard zones associ-
ated with an accidental release Exceptions are possible and likely, however,
so the application of safety factors by users is encouraged
In selection and development of hazard assessment models for ARCHIE ,
particular emphasis was placed on simplicity of computational procedure,
minimization of required input, applicability to a wide range of accident
scenarios, and reasonable accuracy of results Individual subsections below
address individual models in the computer program
aThis appendix was prepared using a special word processing system because of the
large number of equations presented This prevented matching of the fonts and type sizes
used throughout the rest of this handbook
B-l
-------
B.2 Liquid Discharge Models
For a given liquid height and vapor space pressure, the instantaneous liq-
uid release rate from a tank or other container is commonly given by the
equation-
[2gp, (HL -Hh) + 2 (P0 - Pa)] (B 1)
where
m = Discharge rate, kg/s
g — Gravitational constant, 9 8 m/s2
pi = Liquid density, kg/m3
P0 = Storage pressure, N/m?
Pa = Ambient pressure, N/m?
— Liquid height above bottom of container, m
= Height of discharge opening, m
Ah = Area of discharge opening, m2
Ca = Discharge coefficient
An average release rate may be obtained for tanks operating at atmospheric
pressures by computing the time (Te) required to empty a tank Closed form
expressions are available for storage tanks of specific geometries [27] and are
as follows for the system of units being utilized
• Rectangular Tank
_ S
Te ~
Vertical Cylindrical Tank
.. /9 w™ nl
(B3)
B-2
-------
• Horizontal Cylindrical Tank
• Spherical Tank
16 D*T*
Te = 15 (B 5)
where
Tank diameter, m
D0 ~ Opening diameter, m
HT — Liquid height, m
Te = Time to empty, sec
In deriving the above equations, it is assumed that the tank is full and a
circular opening of diameter D0 is at the bottom of the tank The average
release rate is therefore given by
M /13 ^
mavg = jT (B 6J
where M is the total liquid mass in the tank in kilograms
Equation B 1 is also used to calculate pressurized liquid release rates In
this case the difference between the tank pressure and the ambient pressure
is much larger than the liquid head When assuming that liquid head is
negligible, Equation B 1 becomes
m = AhCd^pi2(P0-Pa) (B 7)
This model should be used to calculate liquid discharge from vessels where
the discharge outlet is less then 4 inches from the inner wall of the tank
It has been shown by Fletcher [18,19] that a length of pipe of 4 inches is
necessary and sufficient to establish two-phase flow independent of the pipe
diameter size The more appropriate discharge rate equation for this latter
condition is presented in section B 6
B-3
-------
B.3 Gas Discharge From a Tank
This model calculates the initial discharge rate of a gas from a pressure
vessel. The model assumes that the process is adiabatic and that wall friction
effects are negligible Using the mechanical energy balance, an expression
for an instantaneous discharge rate under non-choked flow conditions may
be calculated from
(B8)
Under choked flow conditions, the mass flow rate is calculated from
l\
where-
m = Discharge rate, kg/s
Ah = Opening area, m2
7 = Ratio of specific heats
Po = Tank pressure, Pascals
pi = Ambient pressure, Pascals
Po = Density, kg/m3
It can be shown that choked flow (maximum flow rate) occurs at a critical
pressure ratio of
2 \^i
Ti) (B10)
These equations are based on ideal gas behavior They can be modified
using a compressibility factor correlation such as Pitzer's correlation [46]
but are not so in ARCHIE for model simplification purposes
B-4
-------
B.4 Gas Discharge From a Pipeline
Two primary computational appi caches aie available for estimation of gas
discharge rates from ptmctuied or ruptuied pipelines conveying strictly
gaseous products The pipeline can be considered to be a volume of com-
pressed non-flowing gas or it can be consideied as a length of pipe with gas
velocity increasing towards the discharge location
The volume model is simple and relies upon the equations described earlier
for gas discharge from a tank It essentially neglects the effects of friction
along the pipe and therefore provides a conservative estimate of the discharge
rate This is the model used in ARCHIE
Consideration of fnctional resistance to flow often requnes graphical tech-
niques to calculate discharge lates The charts refeired to are those de-
scribed by Lapple [35] and later coriected by Levenspiel [37] It is also
possible to use a complex numerical algonthm to solve the mass and energy
conservation equations needed to describe the problem, with the flow being
divided into two parts
1 A fnctionless adiabatic flow at the junction of the leservoir and the
pipe
2 An adiabatic expansion with friction thiough the pipe
Weiss et al [51] present a comparison between field test data, volume model
results, and results obtained using a complex numerical model for pipeline
blowdown calculation Although the volume model underpredicts the blow-
down time, it leads to generally conservative vapoi dispeision calculations
since it overpredicts the dischaige rate foi any given amount of gas in a
pipeline (See Figure B 1)
B-5
-------
Figure B 1- Models vs Field Measurements Taken from Weiss et al
1E+00
%. 1E-01
Q.
1E-02
s X
Vv
^'VvX
^*"0
till
V
^
/
/
a
i i i i
-------
g = Gravitational acceleration, m/s2
gc = Newton's law proportionality factor, 1 0
Va = Inlet velocity, m/s
Vj = Outlet velocity, m/s
B.6 Two-Phase Flow Prom a Tank
This method is used to estimate the two-phase gas-liquid discharge rate
from a tank or other vessel when the discharge outlet is more than four
inches from the inner wall of the vessel and the vessel contains a liquefied
compressed gas
The model is based on theoretical and experimental studies conducted by
Fauske and Associates [15] under the sponsorship of the Design Institute
for Emergency Relief Systems (DIERS) Details of the model are given in
Fauske [15] and Lueng [38]
The final equation describing flashing two-phase flow through a line greater
than 4 inches in length is
ft -010*1* (BIS)
Where
DT — discharge rate, Ibs/minute
D0 = line or pipe diameter, inches
T = absolute temperature, degrees R
Cp = liquid specific heat, jf^
gy = slope of vapor pressure curve,
B-7
-------
Equation B 12 is derived from an equilibrium critical flow model Fauske
(1985) has determined that for a pure flashing component exiting through a
line greater than 4 inches in length, the flow rate may be approximated by
G =
dP,
s
Tgc
(B13)
dT
where
G = mass flux
-J£ = slope of vapor pressure curve
T = absolute temperature
Cp — liquid specific heat
-------
T& = boiling point (the vapor pressure is 760 mm of mercury)
The heat capacity of liquid is not a parameter given in material safety data
sheets, but the range of heat capacities for various types of materials is
relatively narrow
• Organic materials made up of predominantly carbon, hydrogen, oxy-
gen, nitrogen and sulfur typically have a heat capacity ranging from
0 3 Btu/lb-F to 0 8 Btu/lb-F
• Chemicals containing chlorine, fluorine and silicon typically have heat
capacities in the range of 0 2 to 0 4 Btu/lb-F
• Chemicals containing bromine and iodine and metals containing or-
ganics typically have heat capacities ranging from 0 1 to 0 2 Btu/lb-F
In the absence of specific data for a chemical, the program and Chapter 12
suggest use of the following values
• Chemicals containing carbon, hydrogen, oxygen, nitrogen and sulfur
Cp = 0 ZBtu/lb - F
• Chemicals containing chlorine, fluorine or silicon Cp = 0 2Btu/lb— F
• Chemicals containing iodine, bromine or metals Cp = 0 IBtu/lb — F
Equation B 12 can be utilized by first applying Equation B 14 to determine
the approximate slope of the vapor pressure curve and then identifying or
selecting an appropriate liquid heat capacity
B.7 Pool Size Estimation Methods
Estimation of pool areas resulting from discharges of liquid into the ter-
restrial environment is one of the most difficult and error prone aspects of
B-9
-------
accident scenario evaluations for hazardous matenals Exceptions only in-
volve those cases in which the discharge source is confined by a secondary
containment system of known dimensions and the liquid can be expected to
cover the exposed surface of the containment area In the real world, uncon-
fined spills rarely occur in a location where the giound surface if perfectly
flat and impermeable Rather, spilled liquids will typically follow rainwater
discharge paths while simultaneously vaporizing, burning, and / or soaking
into the ground Thus, this model is actually comprised of a number of
different estimation procedures, all of which aie designed to ease the task
of the program user in obtaining a result that is reasonable and likely to be
within the correct ballpark
In the case of a liquid not expected to boil upon iclease to the environment
due to the relationship between its noimal boiling point and ambient and
storage temperatures, ARCHIE first estimates the rate at which the liquid
will evaporate on a unit area basis using a conelation developed by the U S
Air Force that is described in section B 8 This evapoiation flux is then used
as input to a generalized pool spieading model described in section B 10,
with substitution of the evaporation regression late for the burning velocity
of the liquid The result is a maximum credible area for the scenario being
evaluated under specified environmental conditions
Remaining pool size estimation techniques for non-boiling liquids are simple
and permit refinement of results A pool area may be calculated based on
user supplied data from observations at the potential accident or incident
site, the user may select use of the maximum ci edible pool area, or the
user may choose use of a simple and veiy ciude conelation based on limited
experimental data, this being
log(A) = 0 492 log(m) + 1617 (B 15)
where
B-10
-------
m = Total liquid mass spilled, Ibs
A = Pool area, ft2
Computation of pool areas for boiling liquids is accomplished via use of the
boiling rate models described in section B 8 to determine the vaporization
flux and the same pool spreading model described in section B 10 In this
latter case, the vaporization flux is substituted for the burning velocity of
the liquid in appropriate units
B.8 Emission Rates From Liquid Pools
Pools of evaporating liquids are modeled with one of two methods, depending
upon the volatility, normal boiling point, and storage temperature of the
liquid and their relationship to the ambient temperature [11,14,31]
If a discharged liquid is near ambient tempeiatuie, a simplified model de-
veloped by the U S Air Foice Engineering and Seivices Laboiatory [32]
is used to predict the evapoiation late This model was correlated to a
complex numerical model [31], also developed by the Air Force and vali-
dated with experimental data [11] It requires far less user input, yet yields
reasonably accurate answers The model applies to spills that can vary up
and down in temperature as the liquid pool heats and cools, and is limited
in emission rate by how fast mass can transfer into the air from the pool
surface It is also reasonably accuiate foi pools being heated moderately by
the sun, but does not fully account for the decreases in evaporation rates
that boiling or very volatile liquids may expenence as a result of cooling by
evaporation However, the error applies only to one specialized category of
spills, and normally results in a conseivative answer Inaccuracies resulting
from heating of a relatively low volatility substance spilled onto a hot surface
or warmed by the sun over time are counterbalanced by recommendations
B-ll
-------
made to users of ARCHIE in Chapter 12 with respect to specification of
ambient and liquid storage temperatures
The equation used to calculate the evaporation flux of volatile liquid is
Ev = 4 66 x 10-6^°75rFJyfu) (B 16)
where
Ev = Evaporation flux, Ibs/mtn/ft2
Uw — Wind speed, mph
Pa = Vapor pressure of chemical, mm Eg
Pah = Vapor pressure of hydrazme, mm Eg
Mw *= Molecular weight of chemical
TF — Spill temperature correction factor
The spill temperature correction factor is defined as follows.
TF = 1 Tp < 0 C (B 17)
TF = 1 + 4 3 x l(T3rp2 Tp > 0 C (B 18)
where Tp is the pool temperature in degrees C The vapor pressure of hy-
drazine is given by the following equation
7245 2
ln(P) = 65 3319 8 221n(T) + 6 1557 x 1(T3 T (B 19)
where T is in kelvins and P is an atmospheres Note that the total vaporiza-
tion rate of the pool is obtained by multiplying the evaporation flux by the
pool area The duration of vapor emission is obtained by dividing the total
mass of discharged liquid by the total vaporization rate Equation B 18 is
used by ARCHIE for all non-boihng liquids and for those liquids which
boil at pool temperatures in excess of zero degrees Celcius
B-12
-------
The vaporization rate of cold boiling liquids, including most liquefied gases,
is normally driven by the rate of heat transferred from the ground by con-
duction Accurate computation of vaporization rates by so-called ground
conduction models requires knowledge of several ground surface properties
as well as the physio-chemical properties of the spilled material Further
complicating the proper use of such models is the fact that rates will vary
with time as the surface beneath the pool is cooled Notwithstanding the
above, several existing models without excessive data demands were tested
for inclusion in ARCHIE , including a novel approach based on observa-
tions of the relationship between the boiling and burning rates of liquids
which provided the best overall results
An understanding of the logic applied during development of the latter ap-
proach requires knowledge of the following observations
1 Larger differences in temperature between the ground and the boil-
ing point of the discharged liquid lead to higher vaporization rates in
general
2 The depletion rate of a boiling pool at low temperatures can approach
but not exceed the expected rate of depletion if the pool is burning
3 Burning rates of most common flammable liquids vary within a rela-
tively narrow range regardless of the boiling point of the liquid
4 The burning rate of a liquid is a function of its boiling point, molecular
weight, and density
Based on the above findings, a simplified method was developed by correlat-
ing burning velocities estimated by the equation presented in section B 10
with experimentally derived boiling rates for a variety of hazardous materi-
als, these including butane, sulfur dioxide, propane, methane, and oxygen
B-13
-------
The resulting correlation was
F = 0 5322 - 0 001035T6 (B 20)
Ev = Fyp (B21)
where,
Ev = Vaporization flux, kg/m?/s
Tb = Boiling point, F
p = Liquid density, kg/m3
y — Burning velocity, m/s
This approach provides answers of reasonable accuracy within the correct
order of magnitude It is slated for further refinement in a future varsion of
ARCHIE .
B.9 Vapor Dispersion Model
The size of a dispersion hazard zone depends upon the quantity of the ma-
terial released, its effective density, volatilization, prevailing atmospheric
conditions, source elevation, and the user specified toxicity limit
The toxicity limit must be selected by the user carefully to reflect both
the impact of interest (fatality, serious injury, injury, etc ) and the sce-
nario release conditions (especially duration of release or pool evaporation)
Chapter 6 of the guide discusses this topic in detail
The model in ARCHIE is used to determine the downwind distances where
the concentrations are at or above a user specified toxic limiting concentra-
tion. Among the models required for hazard assessments, vapor dispersion
B-14
-------
models are perhaps the most complex This is primarily due to the varied
nature of release scenarios as well as the varied nature of chemicals that may
be released into the environment
In general, dispersion of toxic gases or vapors is influenced by the following
parameters
• Release Rate and Duration
• Prevailing Atmospheric Conditions
• Limiting Concentration
• Duration of Release
• Elevation of the Source
• Surrounding Terrain
• Source Geometry
• Initial Density of the Release
Each one of these parameters are discussed with special emphasis on their
influence on estimation of downwind distances
Release Quantity or Release Rate refers to the total amount of hazardous
chemical which has the potential to be released over a given period of time
in the event of an accident This parameter is a major factor in determining
the dispersion distance In general, larger quantities lead to larger dispersion
distances However, the dispersion distance does not increase linearly with
quantity or release rate In fact, a factor of 10 increase in release rate
usually increases the dispersion distance by a factor of about 3 A factor
of 100 increase in release rate may lead to a factor of 10 increase in the
dispersion distance For gaseous and high vapor pressure liquid releases, the
release rate to the atmosphere will be the same as the discharge rate from
B-15
-------
a vessel or pipeline For liquids with low vapor pressures, the vapor release
rate will be governed by the liquid-specific evaporation characteristics, the
spill area, and the ambient conditions; it can never exceed the liquid spill
rate
Prevailing Atmospheric Conditions include a representative wind speed and
an atmospheric stability class Less stable atmospheric conditions result
in shorter dispersion distances than more stable weather conditions Wind
speed affects the dispersion distance inversely Since weather conditions at
the time of an accident can not be determined a pnon, it is usually prudent
to exercise the model for at least typical and worst case weather conditions
for hazard analysis purposes
Limiting Concentration affects the dispersion distance inversely Lower con-
centrations lead to larger dispersion distances As with source release rate,
the effect is not linear, with a factor of 100 reduction in the limiting con-
centration resulting in an increase in the dispersion distance by a factor of
about 10
Duration of Release is a parameter dependent on release conditions Most
dispersion models use one of the two extreme cases, i e , continuous release or
instantaneous release In the case of instantaneous release (i e , for very short
duration releases), the total quantity of the chemical released during the
accident contributes to the dispersion hazard, and the dispersion takes place
in longitudinal (along wind), lateral (across wind) and vertical directions In
case of a continuous release (i e , a release that lasts a long time compared
with downwind travel time), the release rate is the important parameter,
and dispersion is commonly assumed to take place only in the lateral and
vertical directions
Clearly, most releases do not fall into either one of the above two categories.
Models developed to predict strictly continuous or instantaneous releases
B-16
-------
cannot be applied with reasonable accuracy because they do not take into
account the actual release duration Consequently, a finite duration cor-
rection (see Palazzi [42]) is incorporated into ARCHIE This model ap-
proaches the two limiting cases as the release duration is varied from very
short to very long times The finite duration correction was validated by
Palazzi et al using the experimental data of De Faven et al [13]
The concentration at any location is given in ARCHIE by
(B22)
_ c ,
Cf = -
x - uw (t - tR}
where
Cc =
x
(B23)
exp --
2
-------
x = Downwind distance, ra
H = Source height, m
ax = Longitudinal standard deviation, m
cry = Lateral standard deviation, m
oz = Vertical standard deviation, m
The model uses Pasquill-Gifford [43] dispersion coefficients (i e , standard
deviations) which provide a measure of the turbulence intensity in the lat-
eral and vertical directions These coefficients are a function of downwind
distance It is customary to assume for short duration sources that the longi-
tudinal dispersion coefficient is identical to the lateral dispersion coefficient
As the duration of the release becomes small, the results approach that of
an instantaneous model, conversely, as the duration becomes very large, the
results resemble that of a continuous point source model
Elevation of the Source is attributed to its physical height (such as a tall
stack). In general, the eifect of source height is to increase dispersion in
the vertical direction (since it is not ground restricted), and reduce the
concentration at ground level
Surrounding Terrain affects the dispersion process greatly For example,
rough terrains involving trees, shrubs, buildings and structures usually en-
hance dispersion, and lead to a shorter dispersion distance than predicted
using a flat terrain model Building and terrain effects are site-specific and
cannot be considered in a generalized dispersion model
Source Geometry refers to the actual size and geometry of the source emis-
sion. For example a release from a safety valve may be modeled as a point
source. However, an evaporating pool may be very large in area and may
require an area source model The source geometry effects are significant
B-18
-------
when considering near-field dispersion (less than ten times the character-
istic dimensions of the source) At farther distances, the source geometry
effects are smaller and eventually become negligible Thus, ARCHIE uses
a point source model in all cases Since most toxic substances have long dis-
persion distances, this level of detail was considered adequate for emergency
planning purposes
Initial Density of the release affects the dispersion process A buoyant release
may increase the effective height of the source By the same token, a heavier
than air release will slump towards the ground For heavier-than-air releases
at or near ground level, the initial density determines the initial spreading
rate. This is particularly true for large releases of liquefied or pressurized
chemicals where flashing of vapor and formation of liquid aerosols contribute
very significantly to the initial effective vapor density and therefore to the
density difference with air
Results of recent research programs dramatically indicate the importance
of heavy gas dispersion in the area of chemical hazard assessment In fact,
heavy gas dispersion phenomena exhibit a predictable pattern
• The initial rate of spreading (often termed slumping) is significant
and is dependent on the differences between the effective mean vapor
density and the air density
• The rapid mixing with ambient air due to slumping leads to lower
concentrations at shorter distances than those predicted using neutral
density dispersion models
• There is very little mixing in the vertical direction, and thus, a vapor
cloud hugging the ground is generated
• When the mean density difference becomes small, the subsequent dis-
persion is governed by prevailing atmospheric conditions
B-19
-------
Since heavy gas dispersion occurs near the release, it is particularly impor-
tant when considering large releases of pressurized flammable chemicals To
examine the feasibility of having a simplified version for ARCHIE , the
results of a detailed heavy gas model [41] were compared with those from
a neutral buoyancy vapor dispersion model for both flammable and toxic
hazard scenarios The models were exercised for the following two scenarios
with both neutral and stable atmospheric conditions
• Liquefied propane gas (LPG) released instantaneously with 35 % flash-
ing into vapor and the remaining liquid in the form of aerosols within
the vapor cloud The mass of release was varied from 1 ton to 100
tons The limiting concentration was chosen to be one-half of lower
flammabihty limit for propane
• Pressurized releases of chlorine with 20 % flashing and the remaining
80 % in the form of liquid aerosols The mass of release varied from
0 1 ton to 10 tons The limiting concentration was chosen to be the
IDLH for chlorine (25 ppm)
The results of the model comparison were as follows
• The heavy gas dispersion model, when exercised for the LPG scenario,
gave very similar dispersion distances for both stability categories The
typical distances were between 1000 ft (1 ton release) to 3000 ft (100
ton release).
• The neutral buoyancy dispersion model, when applied to the LPG sce-
nario produced to larger dispersion distances under stable weather con-
ditions The dispersion distances were typically 50 % to 100 % greater
than heavy gas dispersion model predictions for neutral weather condi-
tions For stable weather conditions, the neutral buoyancy dispersion
distances were about 3 to 5 times greater than those given by the
B-20
-------
heavy gas dispersion model
• The heavy gas dispersion model results for chlorine release scenar-
ios were sensitive to the atmospheric stability conditions with stable
weather leading to larger dispersion distances than neutral weather
conditions The heavy gas dispersion model resulted in slightly larger
dispersion distances than the neutral buoyancy dispersion distances
largely due to limited vertical mixing Over the range of parameters,
the differences between the two model results were about 10 %
Based on this comparison, it was concluded that use of a neutral buoy-
ancy dispersion model would be adequate for toxic release scenarios being
evaluated by ARCHIE for emergency planning purposes
The time of arrival of a cloud or plume at a downwind point is often calcu-
lated by simply dividing the downwind distance by an average wind speed
''arrive
TJ (B 26)
t/ty
where
Xdw = Downwind distance, m
Uw = Average wind speed, m/s
In the above expression, the wind speed is assumed to be constant and
independent of height The vapor cloud is assumed to travel at a velocity
similar to that of ambient air
A more realistic estimate may be calculated using a wind speed profile with
a power law expression
1014
—) (B 27)
where
B-21
-------
Uref = Reference wind speed, m/s
zreif = Reference height, m (usually 10 m)
To provide conservative estimates of cloud departure and arrival times,
ARCHIE uses the following logic
• For discharges with source elevations less than 10 meters, the wind
speed reported at a 10 meter height is used to calculate arrival time,
50 % of this speed is used to calculate departure time
• For discharges with source elevations greater than 10 meters, the power
law is used to calculate wind speed for arrival time, 50 % of the 10
meter reference wind speed is used to calculate departure time
Recommended initial evacuation zone widths are estimated using the method-
ology described in Chapter 3 of this guide
B-22
-------
B.10 Pool Fire Model
Upon ignition, a spilled liquid hydrocarbon pool will burn m the form of a
large turbulent diffusion flame Calculating the incident flux to an observer
involves four steps geometric characterization of the flame, estimation of
flame radiation properties, estimation of attenuation coefficients, and com-
putation of the geometric view factors between the observer and flame The
size of the flame will depend upon the spill surface and thermo-chemical
properties of the spilled liquid In particular, the diameter of the fire (if
not confined by a dike), the visible height of the flame, and the tilt and
drag of the flame due to wind can be correlated with the burning velocity
of the liquid The radiative output of the flame will depend on the fire size,
the extent of mixing with air and the flame temperature Some fraction of
the thermal radiation is absorbed by carbon dioxide and water vapor in the
intervening atmosphere In addition, large hydrocarbon pool fires produce
thick smoke which can significantly obscure flame radiation Finally, the in-
cident flux at an observer location will depend on the radiation view factor
which is a function of the distance from the flame surface, the observer's
orientation, and the flame geometry
Experimental data on thermal radiation hazards suggest that an incident
flux of about 5 kW/m2 (1600 Btu/hr-ft2) will cause second-degree burn
injuries on bare skin if the duration of exposure is about 45 seconds An
incident flux level of 10 kW/m2 (3200 Btu/hr-ft2) quickly causes third-
degree burns that are likely to lead to fatality These two levels are typically
used in determining injury and fatality hazard zones (See Buettner [10])
Estimating the thermal radiation hazards from pool fires involves three main
steps characterization of the flame geometry, approximation of the radiative
properties of the fire, and calculation of the safe separation distance to
specified levels of thermal radiation The model is used to calculate the
B-23
-------
following parameters
• Fuel burning velocity,
• Effective emissive power, and
• Fatality and injury hazard zones
Several simplifying assumptions have been made in the calculation proce-
dure. These are summarized below
• Pool area is circular
• Observer is at ground level
• Ambient temperature is 20 degrees C
• Atmospheric absorption of thermal radiation is negligible
• Negligible wind in the vicinity of the flame, thus, uniform thermal
radiation field radially and no flame tilt
• When the user assumes pool ignition occurs shortly after release, the
fire achieves a size such that the burning rate equals the spill rate, with
some adjustments made for special circumstances Otherwise, the fire
has the base area determined via the use of the evaporating or boiling
pool area estimation models
The burning velocity of a liquid pool is the rate at which the pool level
decreases with time The mass burning rate is a related term, being a prod-
uct of the burning velocity and the fuel liquid density Extensive burn rate
measurements (See Burgess [9]) have shown a definite relationship between
the burning velocity and thermochemical fuel properties, such as the ratio
of the net heats of combustion and vaporization The single most readily
available property that best correlates with these heats is the normal boiling
point. Therefore, a simple expression for the burning velocity was obtained,
B-24
-------
covering a wide range of boiling points It is important to note that the
correlation developed is independent of pool size In effect, it is assumed
there is a large, turbulent diffusion flame behaving as an optically thick gray
body This condition is satisfied for most pool fires exceeding about 10 feet
in diameter The equation to estimate the burning velocity is 2
(B28)
where
y = Burning velocity, m/s
Mw = Molecular weight, kg/kgmol
p = Liquid specific gravity
TB = Normal boiling point, degrees F
The spectrum of hydrocarbon liquid spill scenarios is wide Spills can be
classified based on the rate of release and duration
• Continuous spills - in which the spill continues at a specified finite rate
for a long duration
• Instantaneous spills - in which all of the spill occurs in a very short
time
• Finite duration spills - where a given volume of liquid is spilled over a
given duration of time Both the release rate and the release duration
are finite
The diameter of the pool fire depends upon the release mode, release quan-
tity (or rate) and the burning rate In addition, if the spill occurs on land,
2This equation undeipredicts the burning rate for hydrogen ARCHIE therefore in-
cludes a special adjustment for hydrogen
B-25
-------
the factional resistance offered by the terrain will limit the spreading ve-
locity of the liquid In the case of a continuous spill, the liquid spreads
and increases the burning area until the total burning rate equals the spill
rate. This condition of equilibrium is represented by an equilibrium diame-
ter given by the following equation
(B29)
v *y
Here,
J}eg = Steady state diameter of the pool, m
V = Liquid spill rate, m3/s
y = Liquid burning rate 3, m/s
This equation assumes that the dominant mode of heat transfer to the liquid
pool comes from the flame and the burning rate is constant This is a
valid assumption for all liquid hydrocarbons whose boiling temperatures are
above ambient This is also true for liquefied hydrocarbon spills on water
where heat transfer from water to the pool is relatively constant This
results in a higher burning rate The equation, however, ignores the time
dependent heat transfer from substrate such as on land where heat transfer
decreases with time It is also assumed in deriving this equation that the
mass balance is maintained within the burning pool Hence, the loss of
liquid due to percolation through the soil or dissolution in the water column
are not included
In the absence of factional resistance during spreading, the equilibrium di-
ameter given by Equation B 29 is reached over a time given by the following
equation
(B30)
3 see section B 5 for equation
B-26
-------
where, A is the effective gravity
The effective gravity is equal to the gravitational constant for spills on land
It is important to note that the equilibrium diameter does not represent the
maximum diameter of the flame The excess volume spilled up to the time
to reach the equilibrium diameter spreads further The maximum diameter
is given by
Ana* = V2Deq (B 31)
If the spill duration is less than teq, a steady state diameter is not reached
The maximum diameter and the time to reach this diameter are given by
the following expression
[
Ana* = I + I Dch (B 32)
(B33)
(B34)
'
Here,
V8 = Volume remaining in the pool at the end of the spill, m3
ts = Spill duration, sec
Da' = Spill diameter at the end of the spill duration, m
A similar expression may be obtained for simultaneous spreading and burn-
ing of instantaneous releases in the absence of fnctional resistance during
spreading Here the radius of the pool increases until all the material is
consumed by the fire The expressions for the maximum diameter and time
to reach maximum diameter are as follows
= 17766 P^ (B35)
L y2 J
B-27
-------
(B36)
Equations B 32, B 33 and B 34 may be used to derive the above equations
by substituting the total liquid volume V in place of Va and setting the
initial diameter Ds and spill duration ts to zero
For liquid hydrocarbon spills on land, the spreading velocity is largely con-
trolled by fnctional resistance offered by the terrain The maximum pool
diameter and the time to reach the maximum for an instantaneous release
are given by the following expressions
= 17W2
tmax = 05249 ^ (B38)
It should be noted that an instantaneous unconfined pool fire grows in size
until a barrier is reached or until all the fuel is consumed Therefore, the
maximum diameters predicted by these equations will exist only for a short
duration Use of maximum pool diameter will therefore lead to very conser-
vative results A time averaged pool diameter can be obtained by integrating
the time dependent expressions for pool diameter or more appropriately by
dividing the maximum diameter by the square root of two, this being the
approach taken m ARCHIE
A similar expression may be devised for a continuous spill on ground, taking
frictional resistance into consideration The equilibrium diameter is given
by equation B 29 The maximum diameter and time to reach the maximum
diameter are as follows
(B39)
(B40)
B-28
-------
Here Cd is the ground friction coefficient It has been assigned a value of
0 5 in ARCHIE for general application purposes
The criteria by which a given spill situation can be categorized as instan-
taneous or continuous are difficult to establish Comparisons can be made
only between the rapid release of a given volume of liquid and the release of
the same volume of liquid relatively slowly One criterion for classification
is the maximum radius of the burning pool That is, for a given situation,
the maximum radii of spread are calculated using both instantaneous and
continuous models, and the spill is classified into the category which gives
the smaller of the two spread extents Raj [45] indicates that a spill can be
treated as instantaneous if its dimensionless time, T, is less that 0 002
where, ts is the spill duration
The mean visible flame height, estimated by ARCHIE is based on the
correlation of experimental data for laboratory-scale wooden crib fires (See
Thomas [49]) which agrees well with observations of actual liquid pool fires
Based on these experimental data, Thomas developed a correlation for the
mean visible flame height, Hfiame
061
BVp
where
flame = 42Dp
= Flame height, m
p = Liquid density, kg/m3
pa = Air density at ambient temperature, kg/m.3
Dp = Pool diameter, m
g = Gravitational acceleration, 9 8 m/s2
B-29
(B42)
-------
The emissive power of a large turbulent fire is a function of the black body
emissive power and the flame emissivity The black body emissive power, in
turn, can be computed using Planck's law of radiation, if the mean radiation
flame temperature is known For incident flux calculations, however, it is
more important to estimate the effective emissive power of the flame, which
accounts for shielding by surrounding layers of smoke for liquid hydrocarbon
fires Based on observed values of emissive powers reported in the literature
and other available data (See Huggland et al [24] and Alger et al [2]),
the effective emissive power was correlated to the normal boiling point for
selected fuels by the expression
EP = -0 313 TB + 117 (B 43)
where
Ep = Effective emissive power, kW/m?
TB = Normal boiling point in degrees F
Materials with a boiling point above 30 degrees F typically burn with sooty
flames The emissive power from the sooty portion, based on limited data, is
on the order of 20 ^ An effective sooty flame average emissive power can
therefore be estimated by assigning relative areas of sooty and unshielded
flame and calculating an area based average emissive power
The incident flux at any given location is given by the equation
Qmodent = EPXTXVF (B 44)
where
Qtnadent = Incident flux, kW/m?
T = Transmissivity
VF = Geometric view factor
B-30
-------
r, the transmissivity coefficient, is mainly a function of the path- length
(distance from observer to flame surface), relative humidity, and the flame
temperature For the calculation scheme in ARCHIE , r has been set to
1, and the attenuation of thermal flux due to atmospheric absorption is
not taken into account This assumption provides a conservative hazard
estimate, since the presence of water and carbon dioxide tends to reduce the
incident flux at any given location
The view factor defines the fraction of flame that is seen by a given observer
This geometric term has been calculated as a function of distance from the
flame center for an upright flame approximated by a cylinder It has also
been assumed that the optimum orientation between observer and flame
that yields a maximum view factor prevails The resulting equation is as
follows
D -i 1 757
where.
v (B45)
X
X = Distance from flame center, m
Rp — Pool radius, m
For fatality, the incident flux level is set to 10 kW/m2 For injury, the
corresponding level is 5 kW/m2 These levels are based on analysis of nu-
merous sources of experimental burn data (Mudan,1984) Applying these
two damage criteria, the above equations were rearranged to solve for hazard
distances XIQ and XQ$ for fatality and injury, respectively
(B46)
where
= Radius for expected fatalities, feet
B-31
-------
= Radius for expected injuries, feet
B.ll Fireball Model
A large release of a liquefied hydrocarbon (e g , propane) may burn in the
form of a fireball When this occurs, the fireball grows larger and also
moves upwards continuously because of buoyancy The thermal radiation
depends on the size of the fireball, the distance to the observer (which is
continuously changing), and the observer's orientation Experimental data
and observation indicate that the duration of a fireball is typically a few
seconds Therefore, the incident flux at an observer location changes rapidly
with time, and as before, is dependent upon flame shape, emissive power,
attenuation and view factor (See Fay et al [17])
Due to the transient nature of both the size and location of a fireball, the
thermal radiation field also varies with time The equations that follow
define the maximum diameter and height the fireball attains in a short du-
ration, as well as the safe separation distances for fatality and injury Several
simplifying assumptions have been made m the calculation procedure These
are summarized below
• The fuel is propane or has similar characteristics
• Ambient temperature is 20 degrees C (68 degrees Fahrenheit)
• Atmospheric absorption of thermal radiation is negligible
• Fraction of combustion energy radiated = 02
• Observer is at ground level
• Directly under the fireball, the high levels of thermal radiation would
be fatal Therefore, the minimum fatality zone is equal to half the
maximum diameter calculated
B-32
-------
A series of experiments involving pure vapor samples of methane, ethane,
and propane suggest geometric and dynamic scaling relationships for the
geometry of fireballs (Fay & Lewis,1977) The first of these involve the
maximum diameter, height, and duration of the fireball according to the
following equations
Dmax = 16W1/3 (B48)
Z = 263W1/3 (B49)
T = 223Wl/6 (B50)
where-
W = Mass in vessel, Ibs
Dmax — Maximum diameter of fireball, ft
Z = Maximum height of fireball, ft
T = Duration of fireball, seconds
Based on extensive experimental data on pigskin burns, damage criteria for
fatality and injury have been established A Critical Energy Model ,which
assumes that the burn seventy depends upon the amount of energy that is
absorbed by the skin after the surface temperature reaches 55 degrees C,
provides a basis for the damage criteria from fireballs These criteria repre-
sent two levels of thermal radiation a heat flux in excess of 160 kilojoules
per square meter (kJ/m2) could be fatal for humans due to irreversible
skin tissue damage A lesser level of 40 kJ/m2 could cause pain or mild
second-degree burns (Modeled, Wong and Williams, 1973) Therefore, in
this calculation scheme, these two values have been used to define the safe
separation downwind distances from a fireball for fatality and injury, respec-
tively (Mudan & Desgroseilhers,1981) Simple equations were developed by
numerical analysis of results from the rigorous model of fireball hazards
B-33
-------
The safe separation distance for fatality, XF, in feet, was found to be
XF = 148W056 W> 2000/65 (B51)
XF = 8 OW033 W < 2000 /6s (B 52)
Note: The minimum fatality zone equals half the calculated fireball diameter
Dmax jZ & the fireball radius This assumes that anyone directly under the
fireball is in a fatality zone
For injury, the safe separation distance, XI, m feet, is
XI = 4 53W° 52 (B 53)
Note For mass W less than 2000 Ibs, the hazard zones are limited to the
immediate vicinity of the fireball
B.12 Flame Jet Model
The flame length expression for momentum dominated jets is based on mod-
els in the literature with correlating experimental data (See [25,28]) Flame
length is proportional to the molecular weight of the fuel relative to air, and
the diameter of the jet, in addition, it is inversely proportional to the lean
limit concentration of the fuel
For turbulent flame jets, the jet momentum governs the shape of the flame
In addition, the momentum-induced mixing with ambient air is very efficient
and results in higher effective emissive powers for jet flames A detailed flame
jet model takes these differences into account to estimate incident fluxes at
various locations
The flame length correlation by Brzustowski (1973) for momentum- domi-
B-34
-------
nated jets is given by the following equation
Fien 1050
~n—
Djet
where
F[en = Flame length, TO
D3et = jet diameter,m
Cci = Lean limit concentration, vol%
Ma = Molecular weight of air
Mf = Molecular weight of fuel
Some of the theoretical assumptions in this model are as foEows
• For turbulent, momentum-dominated flame jets, flame length is inde-
pendent of mass flow rate
• In order to account for both visible and non-visible portion of flame
length, a concentration criterion is applied Essentially, the end of a
turbulent diffusion flame occurs at that point on the longitudinal axis
where the fuel concentration equals its lean limit In view of extensive
laboratory data, this assumption appears valid
• Effect of crosswmd on flame length not included
• The incident flux levels for observer locations at radial distances from
the flame are not computed Since the direction of the jet cannot be
predicted, it is assumed that the hazards m the longitudinal (axial)
direction represent the maximum distance of impact
• The flame length is correlated against a jet pseudo-diameter for ex-
panded turbulent jets
• The model does not distinguish among vertical, horizontal, or inclined
jet flames
B-35
-------
• Hazards axe evaluated at the plane of the jet For jet flames at an
elevation, the height and orientation of the jet should be considered
The thermal radiation hazards from flame jets have been estimated by char-
acterizing the flame length as a a function of the fuel, and assuming twice
the length dimension to represent the hazard zone
The primary purpose of the model is to point out possible knock-on ef-
fects. The thermal radiation hazards to personnel are generally exceeded by
potential damage to equipment which may be in the path of a flaming jet
B.13 Vapor Cloud Fire Model
This method is applicable for a cloud or plume of flammable vapor which
is released to the atmosphere and is ignited as it drifts downwind A
flame burns and propagates through the space occupied by the cloud The
hazard zone is defined by the length and width dimensions of this cloud
ARCHIE uses the model for toxic vapor dispersion for neutrally buoyant
gases or vapors and a simplified heavy gas model for negatively buoyant
gases to determine the length and width of the hazard zone The heavy gas
procedure was developed as follows
• Using the results of a detailed heavy gas model applied to a number
of pressurized releases, logarithmic plots were developed for instan-
taneous and steady-state releases at neutral (Pasquill D) and stable
(Pasquill F) atmospheric conditions The model was exercised for var-
ious flammable materials, such as propane, ethylene, ethane, butane,
and LNG, over a wide range of release rates and quantities The dis-
tance to the lower flammability limit (LFL) was calculated and corre-
lated with a dimensional parameter, which is a function of the release
B-36
-------
rate (or quantity for instantaneous spill scenarios), and selected prop-
erties of the material, such as molecular weight and LFL The resulting
correlating expressions were found to yield good approximations ol the
more rigorously calculated downwind hazaid zones under a given set
of atmospheric conditions and weie mcoiporated into ARCHIE
• All releases were assumed to occui at or neai groundlevel Elevated
releases of heavy gases are assumed to fall to the ground before dis-
persing
• Ambient temperature was assumed to be 20 C
• In order to estimate the maximum ciosswmd dimensions of the cloud,
the results of the more detailed model weie again reviewed and cor-
related resulting in a series of lules-of- thumb that are applied under
various release conditions
• The effects of finite release duration weie included in a semi-quantitative
manner Given a release rate and duration, the model assumes steady
state, and computes the downwind hazard distance The characteris-
tic travel time of the cloud to this distance is then compared to the
release duration If the release duiation is significantly larger than the
characteristic travel time, then the steady state assumption is consid-
ered valid, if it is smaller, then the lelease is modeled as instantaneous
Where the two terms are comparable in magnitude, finite duiation ef-
fects may be significant In general, the model predicts that, for a
given quantity of flammable matenal, a shoitei release duration re-
sults in a larger hazard zone Since an release assumption provides
the largest zone, relatively shoit-duiation releases were assumed to be
instantaneous
Readers should note that dispersion distances computed using the neu-
trally buoyant vapor dispersion model are generally 2 to 5 times larger
B-37
-------
than those using the heavy gas model Since the width of the flammable
zone typically ranges from 16 to 50 % of the downwind dispersion distance,
ARCHIE assumes 50 % of the downwind hazard as being the approximate
total cloud width for neutrally buoyant gases and vapors
B-38
-------
B.14 Unconfined Vapor Cloud Explosion Model
The explosive effect that may be produced by the ignition of unconfined
flammable vapor clouds is one of the less frequent but severe consequences
of spills Although the precise form of the vapor cloud explosion is still not
fully understood (detonation vs high speed vs low speed deflagration), it is
common practice to express the energy released as a TNT-equivalent charge,
and utilize extensive overpressure data available for TNT explosions (Zab-
etakis [53]) This is accomplished by comparing the combustion energy per
unit mass of a vapor cloud with that of TNT, and taking into account that
only a fraction of the energy in the cloud will contribute to the explosion
Overpressure data compiled from measurements on TNT explosions are then
used to relate overpressure to distance from the explosion Tables B 1 and
B 2 show typical damage criteria as a function of overpressure for all types
of explosions
B-39
-------
TABLE B.I
Explosion Overpressure Damage Estimates
1 Overpressure*
(psig)
0.03
0.04
0.10
0.15
0.30
0.40
0.50-1.0
0.7
1.0
1.0-2.0
1.0-8.0
1.3
2.0
I 2.0-3.0
2.3
2.4-12.2
2.5
3.0
3.0-4.0
4.0
5.0
5.0-7.0
7.0
7.0-8.0
9.0
10.0
15.5-29.0
Expected Damage
Occasional breaking of large windows already under stress.
Loud noise (143 dB); sonic boom glass failures.
Breakage of small windows under strain.
Typical pressure for glass failure.
Some damage to house ceilings; 10% window glass breakage.
Limited minor structural damage.
Windows usually shattered; some window frame damage.
Minor damage to house structures.
Partial Demolition of houses; made uninhabitable.
Corrugated metal panels fail and buckle.
Housing wood panels blown in.
Range for slight to serious injuries due to skin lacerations from flying
glass and other missiles.
Steel frame of clad building slightly distorted.
Partial collapse of walls and roofs of houses.
Non-reinforced concrete or cinder block walls shattered.
1-ower limit of serious structural damage.
Jange for 1-90% eardrum rupture among exposed populations.
50% destruction of home brickwork.
Steel frame building distorted and pulled away from foundation.
'rameless steel panel building ruined.
Cladding of light industrial buildings ruptured.
Wooded utility poles snapped.
Nearly complete destruction of houses.
Beaded train wagons overturned.
1-12 in. thick non-reinforced brick fail by shearing of flexure.
xtaded train box cars demolished.
Jrobable total building destruction.
fcange for 1-99% fatalities among exposed populations due to direct
>last effects.
* These are the peak pressures formed in excess of normal atmospheric pressure by blast and
shock waves.
Source: Lees, RP, Loss Prevention in the Process Industries. Vol. 1, Butterworths, London and
Boston, 1980.
B-40
-------
Reactor: chemical
Filler
Tank- floating roof
Reactor cracking
Utilities gas meter
Juhties* electric transformer
Hlectnc motor
Blower
Fractionauon column
Pressure vessel homofflal
Utilities gas regulator
Extraction column
Steam turbine
Heat exchanger
Tank sphere
Preiiure vessel- vertical
Pump
Overpressure PSI
0.5
A_
A—
10
_c_
1.5
mn_
7,0
«F_
n—
rU
7.5
-T—
"•i" *
10
^f
TM-
K-
15
N
-N
T—
40
-T—
4,5
-B~
-0-
H-
5,0
-T
H-
Q-
5.5
R—
60
-T
-SO
PI-
I-
M
-11
_P_
-II-
70
-.1-
~T
7.5
-1—
I—
1—
80
8.5
90
-T
»I—
LT
-T
9^
-V~l
0,0
— T
-T
•MQ
•>V>»
— — -
110
— T
uJT
-M-
I-
I-
iii' ""
40
*
-T
6.0
-T
-Y
BO
--
20.0
—I
H
—
-
-^
^D~
~V
J
1
wl
I '
A. Windows and quage* break
B. Louven fall at 0.3 - 05 pst
C. Swtchgearis damaged from roof collapse
D. Roof collapses
E. Instruments are damaged
F. Inner parts are damaged
G. Bnckcncki
H. Debris-murile damage occurs
L Unit moves and pipes break
J. Bracing fads
K. Unit uplifts (half-filled)
L. Power lines are severed
M. Controls are damaged
N. Block walls fail
O. Frame collapset
P. Frame deforms
Q. Case is damaged
R. Frame cracks
S. Piping breaks
T. Unit overturns or i» destroyed
U. Una opUfts (09 fined)
V. Una move* on foundations
Source: Stephens, M. M.. Minimiang Damage to Refinenes, VS. Der^ of uielnteriw. Office of Oil &Gts.Febtuary 1970.
TABLE B.2
Explosion Overpressure Damage Estimates
-------
The most serious limitation of the TNT-equivalent model for use in evaluat-
ing this type of scenario is that it may overestimate overpressures in the near
field TNT detonates and produces extremely high pressure shock waves at
short distances with complete destruction of practically everything within
the immediate area Several actual unconfined vapor cloud incidents seem
to indicate, however, that the overpressure within the near field is well below
the level predicted by the TNT equivalent model Based on these data, the
Health and Safety Executive in England (HSE second report [26]), suggests
that the maximum overpressure resulting from an unconfined vapor cloud
explosion model be limited to 1 bar (15 psi) and this recommendation, with
a minor exception, is incorporated into ARCHIE
Of the total energy available for explosion, only a fraction (given by the
yield factor) actually contributes to the explosive effect The yield factor
is probably the most important, yet least precisely known parameter in
explosion hazard analysis It generally ranges from 2 to 20 percent Foi
most aliphatic hydrocarbons, 3 percent is a commonly recommended value
For certain alkenes, approximately 6 percent has been observed Those fuels
that contain oxygen tend to have a higher explosive yields, of up to 16-18
percent Table B 3 presents typical yield factor values for several chemicals
B-42
-------
TABLE B.3
Yield Factors for Explosive Vapors and Gases
Substances With Yield Factors of Y = 03
Acetaldehyde
Acetone
Acrylomtnle
Amyl Alcohol
Benzene
1,3-Butadiene
Butene-1
Carbon Monoxide
Cyanogen
1,1-Dichloroethane
1,2-Dichloroethare
Di-Methyl Ether
Dimethyl Sulphide
Ethane
Ethanol
Ethyl Acetate
Ethylamine
Ethyl Benzene
Ethyl Chloride
Ethyl Cyclohexane
Ethyl Formate
Ethyl Propnonate
Furfural Alcohol
Hydrocyanic Acid
Hydrogen
Hydrogen Sulphide
Iso-Butyl Alcohol
Isobutylene
Iso-Octane
Iso-Propyl Alcohol
Methalamine
Methane
Methanol
Methyl Acetate
3-Methyl-Butene-l
Methyl-Butyl-Ketone
Methyl Chlonde
Methyl-Ethyl-Ketone
Methyl Formate
Methyl Mercaptan
Methyl-Propyl-Ketone
Monochlorobenzene
N-Amyl Acetate
Naphthalene
N-Butane
N-Butyl Acetate
N-Decane
N-Heptane
N-Hexane
N-Pentane
N-Propanol
N-Propyl Acetate
O-Dichlorobenzene
P-Cymene
Petroleum Ether
Phthalic Anhydnde
Propane
Propnonaldehyde
Propylene
Propylene Dichlonde
P-Xylene
Styrene
Tetrafluroethylene
Toluene
Vinyl Acetate
Vinyl Chlonde
Vinyhdene Chlonde
Water Gas
Substances With Yield Factors of Y = 06
Acrolem
Elhylene
Carbon Disulphtde
Etliyl Nitrite
Cyilohexane
Methyl-Vinyl-Ether
Di-Ethyl Ether
Phthalic Anhydride
Pi-Vinyl Ether
Propylene Oxide
Substances With Yield Factors of Y = 19
Acetylene
Isopropyl Nitrate
Ethylene Oxide
Methyl Acetylene
Ethyl Nitrate
Nitromethane
Hydrazine
Vinyl Acetylene
B-43
-------
Although overpressures m the neai field may be significantly ovei estimated,
estimates of overpressure expenenced in the fai field are believed to be more
accurate and applicable Note that the fai field legion is usually of greater
interest because a relatively small piessuie rise is sufficient to cause failures
of buildings and structures
It should be noted that the model in ARCHIE does not account for unig-
mted cloud drift For the total explosion hazaid zone, a drift factor, calcu-
lated using the vapoi cloud file model, must be added to the ladial over-
pressure distance This explains the note to useis in Chaptei 12 that the
center of the equation can be anywheie within the hazaid zone computed
by the vapor cloud fiie model foi the LFL boundaiy
Note that the model gives the user a choice of computing radial hazard zone
distances for a free sphencal ail explosion (above ground) or gioundlevel
explosion (hemi-sphencal) The distances calculated by ARCHIE for fiee
air events are lesser by a factoi of 1 260 than gioundlevel hazaid zone dis-
tances
Selected simplifying assumptions made in the calculation proceduie are
• Ambient temperatuie is 20 degiees C (68 degiees Fahrenheit)
« The mass of explosive is compaied against an equivalent charge of
TNT
• The effects of tenam, buildings, obstacles have not been consideied
The calculational proceduie itself begins by computing the adjusted equiv-
alent mass of TNT using the following equation
= \mcioud X X Yf \ (E 55)
where
B-44
-------
= TNT equivalent mass, Ibs
AHC = Lower heat of combustion, kcal/kg
™>doud = Mass in cloud, Ibs
Yj = Yield factor
The distance to a given overpressure is then calculated from the equation
X = rojfir exp (3 5031 - 0 7241 ln(Op) + 0 0398(ln OP)2) (B 56)
where
X = Distance to given overpressure, ft
Op = Peak overpressure, psi
Data in Lee's [36] which provide the curve of overpressure versus scaled
distance were utilized to obtain these working equations using regression
analysis
B.15 Tank Overpressurization Explosion Model
As discussed in Chapter 5, explosion hazard zones can also result from the
violent rupture of an overpressunzed container without combustion taking
place The energy that produces the overpressure field comes from the vol-
ume and internal pressure of the vapor space in the container Thus, a nearly
empty pressurized container of gas can be a more severe explosion hazard
than a container nearly full of liquid that ruptures at the same pressure
ARCHIE utilizes a model that assumes that the tank ruptures in manner
in which the blast wave propagates omm-directionally (See Baker [3]) It
assumes a ground level, herm-spherical field and yields ground level hazard
zones If the ruptured container is elevated, calculated hazard zone distances
B-45
-------
should be reduced by a factor of 1 26 to yield the radial hazard zone from
the center of the explosion
The computational algorithm proceeds as follows
1. Calculate the ratio Pt/Pa where Pa is the absolute ambient pressure
and P, is the absolute internal gas pressure at which the tank is ex-
pected to rupture
2 Compute the ratio Tt/Ta where Ta is the absolute ambient air temper-
ature and T, is the absolute temperature of the gas in the tank
3 Determine the initial overpressure ratio, Pso, by solving the following
equation by trial and error
/ = o
= ln(Pt/P«)-
p
- —^-ln 1-
1-7.
(7. ~
0 57a(7o
where, 7 is the ratio of specific heat at constant pressure to that at
constant volume
4. Compute the nondimensional starting distance R0 from
1
RO —
5 compute the value of R from
R =
*£(£._-,
3
•y.-l
1/3
(B57)
Kfe-0
1/3
(B58)
where,
B-46
-------
V = Volume of the gas in the tank
r = Distance from the center of the tank at which the side-on over-
pressure is desired 4
6 Locate the point associated with Pso and R0 on Figure B 2 This is
the starting point 5
7 Follow the nearest curve for Ps vs R to the R value computed in step
5 Read the Ps value associated with this R value If the gas vessel is
on the ground and/or close to a reflecting surface, increase Ps by 100
% for R less than 1 and by 10 % for R greater than 1
8 The side-on overpressure is determined by multiplying the above re-
sultant value of Ps by the absolute value ambient atmospheric pressure
Pa
If the volume of the gas in the bursting tank is better represented by a
cylinder 6 than a sphere, adjust the above result for side-on overpressure as
follows
• For R less than 0 3, the calculated overpressure should be multiplied
by a factor of 4 or 5
• For R near 1 0, multiply the answer by a factor of 1 6
• for R greater than about 3 5, multiply the answer by a factor of 1 4
Note that the difference between spherical and cylindrical vessel bursts is
only known qualitatively Therefore, the corrections are very crude
*The above equation is for gas vessel bursts far from reflecting surfaces If the gas
vessel is on ground, multiply V above by a factor of 2, as is done in ARCHIE
SARCHIE uses a curve fit of this graph
6Although not discussed with respect to this estimation procedure, NASA workbooks
typically assume cylinders have a length to diameter ratio of 10 while spheres have a ratio
of one
B-47
-------
Figure B.2: Ps vs. R for Overpressure Calculations. Taken from Baker
10'
10
B-48
-------
B.I6 Condensed Phase Explosion Model
This model calculates the explosion hazard zones resulting from the deto-
nation of solid or liquid explosives This method is essentially the same as
the TNT model upon which the previously described UVCE model is based
This model takes into account the difference in available combustion energy
between the material of concern and TNT Since the explosive material will
also detonate, the yield factor is unity and no limit is placed on the near
field overpressures
B-49
-------
Bibliography
[1] Abramovich, G N (1963), The Theory of Turbulent Jets, English trans-
lation of Russian Publication, MIT Press
[2] Alger, R S , R C Corlett, A S Gordon, and F A Williams (1979),
Some aspects of Twbulent Pool Fires, J Fire Tech , 15, 2, pp 142-156
[3] Baker, W E , (1973), Explosions m Air, University of Texas Press,
Austin, Texas
[4] Erode, H L (1959), Blast Wave From a Spherical Charge, The Physics
of Fluids, Vol 2, No 2, p 217
[5] Bnggs, G A , S R Hanna, and R P Hosker (1982), Handbook on
Atmospheric Diffusion, U S Department of Energy, DOE/TIC-11223
[6] Brzustowski, T A (1971), Predicting Radiant Heating from Flares,
Esso Engineering Research and Development Report, EE 15ER 71
[7] Brzustowski, T A (1973), A New Criterion for the Length of a Gaseous
Diffusion Flame, Comb Sci & Tech , Vol 6, pp 313-319
[8] Burgess, D S and M Hertzberg (1974), Radiation from Pool Fires, m
Heat Transfer in Flame, edited by Afgan, N H and J R Beer, Scripta
Book, Co, Washington
[9] Burgess, D S , A Strasser, and J Grumer (1961), Diffusive Burning
of Liquid Fuel m Open Trays, The Fire Research Abs & Review, 3, p
177
[10] Buettner, K (1981), Effects of Extreme Heat and Cold on Human Skin,
II Surface Temperature, Pain and Heat Conductivity in Experiments
with Radiant Heat, J Appl Physics, 3, p 703
[11] Clewell, H J (1983), A Simple Method for Estimating the Source
Strengths of Spills of Toxic Liquids, ESL-TR-83-03
B-50
-------
[12] Cude, A L (1974), Dispersion of Gases Vented to Atmosphere from
Relief Valves, Chem Eng Land, 290, 629
[13] De Faven, D M , E L Henzevack, and B T Delany (1982), Dispersion
from Safety Valves and Other Momentum Emission Sources Intermit-
tent, Atmospheric Environment, Vol 16, pp 1239-1242
[14] Drake, E M and Reid, R C (1975), How LNG Boils on Soils, Hydro-
carbon Processing, 191-194
[15] Fauske, H K (1985), Flashing Flow Some Practical Guidelines for
Emergency Releases, Plant/Operations Progress, Vol 4, No 3, p 132
[16] Fay, J A and Lewis, D H (1976), Unsteady Burning of Unconfined Fuel
Vapor Clouds , 16th Intl Symp on Combustion, Combustion Institute,
Pittsburgh, PA, 1397 if
[17] Fay, J A,G J Desgroseilliers, and D H Lewis (1979), Radiation from
Burning Hydrocarbon Clouds, Comb Sci Tech , 20, pp 141-151
[18] Fletcher, B and A E Johnson (1984), The Discharge of Superheated
Liquids From Pipes, I Chem E Symposium Series 85
[19] Fletcher, B (1983), Flashing Through Orifices and Pipes, AIChE Loss
Prevention Symposium, Denver
[20] Fumarola, G (1983), Determining Safety Zones for Exposure to Flare
Radiation, Int Chem Eng Symposium Series No 82, pp G23-G30
[21] Galant, S (1984), Three Dimensional Parabolic Calculations of Large
Scale Methane Turbulent Diffusion Flames to Predict Radiation Under
Crosswmd Conditions, 20 th Combustion Symposium, pp 532-540
[22] Goldwire, H C (1985), Status Report on the Frenchman Flat Ammonia
Spill Study, AIChE Symposium, August 25-28
[23] Gugan, K (1978), Unconfined Vapor Cloud Explosions, The Institute
of Chemical Engineers, London
[24] Haggland, B and L Person (1976), The Heat Radiation from Petroleum
Fires, FOA Rapport, Forsvarets Forskmngsanstalt, Stockholm
[25] Hawthorne, W R , D S WeddeU and H C Hottel (1949), Mixing and
Combustion on Turbulent Gas Jets, 3rd Int Comb Symposium, pp
266-288
B-51
-------
[26] Health and Safety Executive (1979), Second Report of the Advisory
Committee on Major Hazards
[27] Hicks, T G , (1985), Standard Handbook of Engineering Calculations,
2nd Edition, McGraw-Hill
[28] Hottel, H C and W R Hawthorne (1949), Diffusion in Laminar Flame
Jets, 3rd Int Comb Symposium, Flames and Explosions, pp 254-266
[29] Hottel, H C and A F Sarofim (1967), Radiative Transfer, McGraw
Hill, New York
[30] Hoot, T G , Meroney, R N and Peterka, J A (1973), Wind Tunnel
Tests of Negatively Buoyant Plumes, EPA Grant AP-01186, Colorado
State University Report CER73-74
[31] Hie, G , and Springer, C , (1978), The Evaporation and Dispersion
of Hydrazme Propellants from Ground Spills, CEFDO 712-78-30, AD
A059407
[32] Kahler, J P , Curry, R C and Kandler, R A (1980), Calculating Toxic
Corridors, AWS TR-80-003
[33] Kalghati, G T (1983), The Visible Shape and Size of a Turbulent Hy-
drocarbon Jet Diffusion Flame in a Crosswind, Combustion and Flame,
Vol 52, pp 91-106
[34] Kalghati, G T (1981), Blowout Stability of Gaseous Diffusion Flames
m Still Air, Combustion Science and Technology, Vol 26, pp 233-239
[35] Lapple, C E (1943), Isothermal and Adiabatic Flow of Compressible
Fluids, Trans Am Inst Chem Engrs , Vol 39, pp 385-432
[36] Lees, F (1980), Loss Prevention m the Process Industries, Butterworth
[37] Levenspiel, 0 (1977), The Discharge of Gases from a Reservoir through
a Pipe, AIChE Journal, Vol 23, No 3, pp 402-403
[38] Lueng, J C (1986), A Generalized Correlation for One Component
Homogeneous Equilibrium Flashing Choked Flow, AIChE Journal, Vol
32, No 10, pp 1743-1745
[39] McCabe, W L and J C Smith (1976), Unit Operations of Chemical
Engineering, 3rd Edition, McGraw-Hill Book Company, New York
[40] Mudan, K S (1984), Thermal Radiation Hazards from Hydrocarbon
Pool Fires, Prog Energy Comb Sci, Vol 10, pp 59-80
B-52
-------
[41] Mudan, K S (1983), Gravity Spreading and Turbulent Dispersion of
Pressurized Releases Containing Liquid Aerosols, IUTAM Symposium,
September 1-3, TheHague, The Netherlands
[42] Palazzi, E , M De Faven, G Fumarola, and G Ferraiolo (1982), Dif-
fusion from a Steady Source of Short Duration, Atmospheric Environ-
ment, Vol 16, No 12, pp 2785-2790
[43] Pasquill, F (1974), Atmospheric Diffusion, 2nd Edition, Wiley, New
York
[44] Putnam, A A and C F Speich (1963), A Model Study of the In-
teraction of Multiple Turbulent Diffusion Flames, 9th Int Symp on
Combustion, pp 867
[45] Raj, P K (1979), Models For Cryogenic Liquid Spill Behavior on Land
and Water, Journal of Hazardous Materials
[46] Smith, J M , and H C Van Ness (1987), Introduction to Chemical En-
gineering Thermodynamics, Fourth Edition, McGraw-Hill Book Com-
pany
[47] Steward, F R (1970), Prediction of the Height of Turbulent Diffusion
Buoyant Flames, Comb Sci Tech , Vol 2, pp 203-212
[48] Strehlow, R A and Baker, W E (1976), The Characterization and Eval-
uation of Accidental Explosions, Progress in Energy and Combustion
Science, 2, pp 27-60
[49] Thomas, P H (1863), The Size of Flames from Natural Fires, 9th
International Combustion Symposium, pp 844-859
[50] Turner, B D (1969), Workbook of Atmospheric Dispersion Estimates,
U S HEW Report 999-AP-26
[51] Weiss, M H , K K Botros and W M Jungowski (1988), Simple Method
Predicts Gas-Line Blowdown Times, Oil & Gas Journal, Dec 12, pp
55-58
[52] Yuen, W W and C L Tien (1976), Simple Calculation Scheme for
the Luminous Flame Emissivity, 16th Int Symp on Combustion, The
Combustion Institute, Pittsburgh, PA, pp 1481-1487
[53] Zabetakis, M (1967), Safety with Cryogenic Fluids, Plenum Press
B-53
-------
-------
APPENDIX C
OVERVIEW OF "SHELTER-IN-PLACE" CONCEPTS
Introduction
There are essentially two ways to protect the public from the effects of toxic gas or
vapor discharges into the atmosphere. One of these methods is evacuation and involves
relocation of threatened populations to shelters in safer areas The other involves giving
instructions to people to remain inside their homes or places of business until the danger
passes In other words, it involves telling people to "shelter-m-place."
Evacuation is clearly safer with respect to the specific hazards posed by a toxic gas or
vapor release but has certain limitations and may pose new problems For example, it is
fairly well appreciated that a major evacuation takes time and may not be feasible once large
amounts of toxic gases or vapors have actually entered the atmosphere Indeed, asking
people in the path of a toxic cloud or plume to leave their homes may actually cause greater
harm than good in some cases. Thus, large-scale evacuations in response to toxic gas or
vapor hazards are best considered when.
• There is a strong potential for a toxic discharge, the discharge has not yet
taken place, and there appears to be time available to relocate people.
• The discharge has taken place but people are sufficiently far downwind to
permit time for evacuation.
• People not yet in the direct path of a cloud or plume are threatened by a
future shift in the wind direction.
• The safety hazards of evacuation are outweighed by benefits of the action,
and/or
• Telling people to shelter-in-place might not fully protect them from serious
consequences.
Much has been written and said over the years pertaining to the merits and
disadvantages of evacuation, but little information has been made available on shelter-
ing-in-place, and indeed, far too many people and organizations have come to believe that
sheltering-in-place will provide adequate protection to the public under a wide variety of
circumstances without actually studying the issues involved. It is therefore the purpose of
c-i
-------
this appendix to explain why staying indoors provides some degree of protection, to discuss
the degree of protection that might be expected, and to discuss how best to instruct the
public to shelter-in-place.
Outside Air Entry Into Buildings
If a building or other structure is airtight, i.e., like a sealed box, people inside will be
completely immune from exposure to any toxic gases or vapors outside its walls On the
other hand, if walls on the upwind and downwind sides of the building are missing, these
people will be exposed to the same level of contamination "indoors" as they would be if they
were in the open. It is easy to understand, therefore, that the exposure of people inside a
structure to toxic gases or vapors in the external environment is a function of the
"airtightness" of the building and the rate at which outdoor air passes in and out.
There are essentially three main ways in which air can enter (and exit) a structure, these
being:
• Natural ventilation
• Mechanical ventilation, and
• Infiltration
Natural ventilation refers to entry of outdoor air into a building through open windows
or doors without assistance from fans. Obviously, the more openings in a building, the
greater the rate at which outdoor air can pass through. Occupants can generally control this
rate by opening and closing various windows and doors
Mechanical ventilation refers to the use of fans and other equipment to bnng air into a
building, possibly heat, cool, filter, and/or recirculate it several times, and then exhaust it
from the structure. This type of system is most often seen in office buildings, other
commercial establishments, and factories. As above, occupants usually have considerable
control over the rate of ventilation.
Infiltration is air leakage into a building through cracks and small openings around
windows and doors and through floors and walls The rate at which an* enters a building by
this mechanism depends on the type of building, workmanship and materials applied during
construction, and the condition of the building. Infiltration differs from natural and
mechanical ventilation in the sense that occupants are generally considered to have little
effective control over its rate. The total rate at which outdoor air enters a building, for the
purposes of this discussion, can be considered as being the sum of the three types of
ventilation described above.
C-2
-------
Rates of outdoor air ventilation are typically expressed in units of air changes per hour
(acph), this being the number of building volumes of outdoor air that enter the building or
other structure in the course of an hour For example, if a building has an internal volume of
10,000 cubic feet, and 20,000 cubic feet of outdoor air enters the building each hour, its total
"fresh air" ventilation rate is said to be 2 0 acph If only 5,000 cubic feet of air enter the
building in the same span of time, the rate becomes 0 5 acph
The natural ventilation rate in structures with open windows and doors can vary widely
and depends on the area of openings, the wind speed, the orientation of openings with respect
to the wind direction, and the building volume It could be as little as a fraction of 1 0 acph
or as high as 80-90 acph, and possibly even more if residents do not mind a strong breeze
blowing through the structure on a windy day
Mechanical ventilation rates in office buildings and the like typically range from 4 to
12 acph with certain exceptions During pleasant weather, 90 to 100 percent of this air might
be from the outdoors In very cold or very hot weather, building operators often have the
option to reduce heating or cooling costs by lowering the fresh air entry rate to 5 to 35
percent of the mechanical ventilation rate by recirculatmg (reusing) large volumes of air.
There are a great many factors that influence infiltration rates in homes and other
buildings When outdoor wind velocities are very low and indoor-outdoor temperature
differences are minimal, infiltration rates may be as low as 0.1 acph The average rate m
American homes, however, is on the order of 0.8-0.9 acph, and "leaky" homes may
experience 2 5 acph or so, especially under poor weather conditions with high winds and low
temperatures Available data on other types of building construction are limited but suggest
an average infiltration rate of about 1.0 acph for office-type buildings
Effect of Total Outdoor Air Ventilation Rates on Indoor Exposures
The overall subject of how outdoor air pollutants can affect exposures indoors is rather
complex and is best discussed using two examples, one for the case when a distinct cloud of
airborne contaminants passes a building, and another for the case when the building is
engulfed by a plume of vapor or gas for a prolonged penod of time The cloud is assumed to
pass in a total time penod of 30 minutes, while the plume is assumed to last 10 hours Both
the cloud and plume are assumed to have an average contaminant concentration of 100 ppm
for the duration of their existence at any outdoor location.
Table C 1, developed from a mass balance model of indoor-outdoor pollutant
relationships, presents average indoor contaminant concentrations expected for the cloud
scenario when outdoor air ventilation rates range from 0.1 to 50 acph Important observa-
tions are that:
C-3
-------
TABLE C.1
INDOOR CONCENTRATIONS FOR HYPOTHETICAL CLOUD PASSAGE
Time
(minutes)
0
5
10
15
20
25
30
40
50
60
90
Internal Concentrations (ppm) at Various Air Change Rates (acph)
0.1
acph
0
08
1.7
25
3.3
4.1
49
4.8
47
46
4.4
0.5
acph
0
4.1
8.0
118
154
188
221
203
187
172
134
1.0
acph
0
80
154
22.1
283
341
393
333
282
238
14.5
1.5
acph
0
118
222
313
393
465
528
411
320
249
118
2.0
acph
0
15.4
283
393
487
565
632
453
324
233
86
2.5
acph
0
18.8
34.1
465
565
64.7
71.3
470
310
204
59
5.0
acph
0
34.1
565
71.3
81.1
875
918
399
173
75
06
10.0
acph
0
565
81.1
918
964
984
993
18.8
3.5
0.7
0
50.0
acph
0
98.4
100
100
100
100
100
0
0
0
0
n
Note. The cloud is assumed to have an average concentration of 100 ppm outdoors. Its leading edge
reaches the building at time equals zero. Its trailing edge passes the building at time equals 30
minutes, at which point internal contaminant concentrations begin to drop
-------
Indoor concentrations increase steadily until the point in time that the
discharge ceases and the last of the airborne contamination passes a
building.
Tight buildings or average buildings in highly favorable weather, with air
change rates of 0 1 acph, are expected to expenence a peak indoor
contaminant concentration only 5 percent of the outdoor average after 30
minutes. Due to the lack of ventilation, however, indoor levels will drop
slowly after the cloud has passed
• "Average" homes and buildings under average conditions, with air change
rates of 0 5 to 10 acph, may expenence indoor concentrations on the order
of 20-40 percent of outdoor levels after 30 minutes
• "Leaky" buildings or average buildings exposed to severe weather condi-
tions, with air change rates of 1 5 to 2 5 acph, may expenence 45 to 65
percent of outdoor concentrations in 30 minutes
• Buildings that have open windows or doors or mechanical ventilation
systems bnnging in outdoor air at high rates will expenence contaminant
concentrations close to those expenenced outdoors.
• Peak indoor levels will be lower if the cloud passes in less than 30 minutes
and higher if the cloud requires more than 30 minutes to pass
• There are benefits to be realized by telling people to open then1 windows
and turn on ventilation systems as soon as possible after the danger has
passed This will help "flush out" any contaminants trapped in buildings
Table C 2 demonstrates average indoor contaminant concentrations expected while a
building is exposed to a plume of vapor or gas for 10 hours Outdoor an* ventilation rates on
the table again range from 0 1 to 50 acph Important observations are that*
• Indoor concentrations increase steadily until the point in tome that the
discharge ceases and the last of the airborne contamination passes a
building
• Average buildings under average weather conditions, with outdoor air
ventilation rates of 0 5 to 1 0 acph, will have indoor contaminant concentra-
tions of 40 to 65 percent of those outdoors in an hour, 60 to 85 percent or so
in two hours, and 75 to 95 percent in three hours Buildings with higher
rates of outdoor air entry will require less time to reach these levels
C-5
-------
TABLE C.2
INDOOR CONCENTRATIONS FOR HYPOTHETICAL PLUME EXPOSURE
Internal Concentrations (ppm) at Various Air Change Rates (acph)
Time
(hours)
0
1
2
3
4
5
6
7
8
9
10
11
12
0.1
acph
0
9.5
181
259
330
39.3
45.1
503
551
593
632
572
51.7
0.5
acph
0
39.3
632
77.7
864
918
950
970
982
989
993
602
365
1.0
acph
0
632
864
950
982
993
998
999
100
100
100
36.8
135
1.5
acph
0
111
970
989
99.8
999
100
100
100
100
100
223
50
2.0
acph
0
865
982
998
100
100
100
100
100
100
100
135
1.8
25
acph
0
918
99.3
99.9
100
100
100
100
100
100
100
82
07
5.0
acph
0
993
100
100
100
100
100
100
100
100
100
07
0
10.0
acph
0
100
100
100
100
100
100
100
100
100
100
0
0
50.0
acph
0
100
100
100
100
100
100
100
100
100
100
0
0
n
Note: The plume is assumed to have an average concentration of 100 ppm outdoors. Its leading edge reaches
the building at time equals zero Its trailing edge passes the building at time equals 10 hours, at which
time internal contaminant concentrations begin to drop.
-------
• Sheltenng-in-place may not make sense when discharges are expected to be
prolonged and outdoor concentrations are expected to be harmful
Recommended Shelter-In-Place Instructions
The previous sections have demonstrated the circumstances under which shelter-
ing-in-place can provide some degree of protection from toxic gases and vapors in the
atmosphere and those circumstances under which the practice may not be effective They
have also demonstrated limitations of the practice and shown how minimization of outdoor
air infiltration and/or ventilation rates into buildings is critical. This section draws upon the
information presented above and substantial other data to present a list of suggested
instructions to be given populations asked to shelter-m-place, these being:
1. Close all doors to the outside and close and lock all windows (windows
sometimes seal better when locked)
2. Building superintendents should set all ventilation systems to 100 percent
recirculation so that no outside air is drawn into the structure Where this is
not possible, ventilation systems should be turned off.
3. Turn off all heating systems
4. Turn off all air-conditioners and switch inlets to the "closed" position Seal
any gaps around window type air-conditioners with tape and plastic
sheeting, wax paper, or aluminum wrap
5. Turn off all exhaust fans in kitchens, bathrooms, and other spaces
6. Close all fireplace dampers
7. Close as many internal doors as possible in your home or other building
8. Use tape and plastic food wrapping, wax paper, or aluminum wrap to cover
and seal bathroom exhaust fan grilles, range vents, dryer vents, and other
openings to the outside to the extent possible (including any obvious gaps
around external windows and doors).
9 If the gas or vapor is soluble or even partially soluble in water — hold a wet
cloth or handkerchief over your nose and mouth if the gases start to bother
you For a higher degree of protection, go into the bathroom, close the
door, and turn on the shower in a strong spray to "wash" the air Seal any
C-7
-------
openings to the outside of the bathroom as best as you can. Don't worry
about running out of ah* to breathe. That is highly unlikely in normal
homes and buildings.
10. If an explosion is possible outdoors ~ close drapes, curtains, and shades
over windows. Stay away from external windows to prevent potential
injury from flying glass.
11. Minimize the use of elevators in buildings. These tend to "pump" outdoor
air in and out of a building as they travel up and down.
12. Tune into the Emergency Broadcast System on your radio or television for
further information and guidance.
As a final note, be advised that this appendix mostly deals with the problem of toxic
gases and vapors in the atmosphere. Since these substances can pass through tiny spaces
where dusts and other aerosols hi air may not, sheltering-in-place may provide a much
greater degree of protection when airborne contaminants are m the form of liquid or solid
particles.
Technical Basis for Internal Concentration Estimates
Given the potential importance of this topic and the possible desire of emergency
planning personnel and others to undertake more formal analyses of indoor/outdoor pollutant
relationships in occupied structures, this subsection of the appendix presents the technical
basis for the estimates shown in Tables C.1 and C.2.
Figure C.1 illustrates a structure with a forced air ventilation system With an initial
assumption that all external windows and doors to the structure have been closed, outdoor air
with a contaminant concentration of Cm enters the internal space via infiltration and a
mechanical make-up air system (i.e., a system using one or more fans to force air into the
structure to replace air supplies being exhausted to the outdoors via exhaust fans). Besides
being mechanically forced out, air also exits the structure via exfiltration in a process similar
but opposite to infiltration.
As is the case in most structures with forced air systems, heating and/or cooling costs
are niinimized by recirculating some fraction of the total internal ventilation rate through an
air filter of some type before reintroduction to occupied spaces. For completeness, the
illustration also shows a source of airborne contaminants within the structure itself, though
the actual presence of such a source is unlikely in the case of the vast majority of hazardous
materials that may be released to the external environment.
C-8
-------
'mu = Outdoors
Contaminant
Concentration
Fresh make-up_
air rate ~ mu
Infiltration = Q|
rate
Qx = Exhaust Rate
o
VO
I
^^ (3
p
fAir A
istributionT
«
System |
^
1 -
A QR= Return
^ Air
5 Rate
= Exfiltration Rate
C - Internal contaminant concentration
G = Contaminant generation indoors
V= Internal space volume
TJ = Fractional efficiency of air cleaner
FIGURE C.1
ILLUSTRATION OF VENTILATED STRUCTURE
-------
Table C.3 presents a formal derivation of the equation necessary to evaluate the
buildup of internal contaminant concentration (C) in the structure as a function of elapsed
time (t) when the structure is first exposed to contamination at time "t0". The resulting
equation is then simplified for use with structures in which forced-air ventilation systems are
turned off (or never existed in the first place) in response to emergency shelter-m-place
instructions from public authorities. This equation is only valid up to the instant in time that
the trailing edge of the external contaminant cloud or plume passes the structure and
therefore only applies while contaminant concentrations are increasing indoors If it is
desired to express the natural air infiltration rate in units of air changes per hour (acph), and
this rate is simply shown as Q, the buildup phase equation becomes:
C = Cm[l-exp(-Qt)]
Thus, if external concentration C^ equals 100 ppm, the total ventilation late is one air change
per hour, and the structure has been engulfed in a plume for one-half hour, the internal
contaminant concentration can be estimated as being 39.3 ppm.
Once the trailing edge of a cloud or plume passes a structure, the external environment
will be generally free of contamination, but it will take additional time for contaminants to be
purged from the structure if new sources of fresh air are not introduced Derivation of the
equation for the purging phase of contamination is similar to that shown in Table C 3 until
the final steps. Instead of being equated to zero, the parameter C0 is used to represent the
concentration in the structure at the instant the trailing edge of the cloud or plume passes
Similarly, parameter t0 is equated to the elapsed time of this occurrence, and finally C^ is set
to zero. The resulting equation for the purging phase becomes-
C = C0exp[-Q(t-t0)]
Thus, if the trailing edge of a plume passes a structure in an elapsed time (0 of one-half
hour, the internal contaminant concentration (C0) is 39.3 ppm at this time, the total ventilation
rate is one air change per hour, and it desired to determine the internal concentration (C) at a
total elapsed time of one hour from initial exposure of the structure, the above equation can
be used to predict a concentration of 23.8 ppm.
The model is based upon several assumptions requiring explanation These involve4
• Perfect mixing
• Use of a single compartment
• Negligible deposition or decay of contaminant
• Perfect air balance
• Constant external contaminant concentration
C-10
-------
TABLE C.3
DERIVATION OF MODEL FOR CONTAMINANT
BUILDUP PHASE IN STRUCTURES
Rate of Accumulation = Rate of Generation - Rate of Removal
VdC = Gdt + CMUQidt + CMUQMudt + (1 - rj)CQRdt - CQxdt - CQEXdt - CQRdt
±fdt = r° _ dC _
V Jt0 Jc0 G + CMUQi + CMUQMU + (1 - n}CQR - CQX - CQEX - CQR
G + CW(<2/ + QMU} + C2(rjQR + Qx + QEX} _ r,QR + QX + QEX ,
(t ~ *o)
+ Qx + Qs^ V
Let A = G
A-
C = h-e-v(
n
r - MU(QI + QMU] \. -s.it-t n , ^
- ,n ,n - - 1 - e v(t to>\ when C0 = 0
+ QX + QEX l -I
Note'
!• Qx + QBX = QMU + QI m all cases.
2- QBX = QI when fans are turned off or are not present.
3. G = 0 where there are no internal contaminant sources.
4- QMU = 0 a*"* QR = 0 wnen fans are turned off or are not present.
5. t0 = 0 when time (t) is measured from the instant a cloud or plume reaches the
structure.
6. C0, the internal contaminant concentration at time = 0, can be assumed to be zero
when there are no internal contaminant sources. With these simplifications:
C =
C-ll
-------
The first assumption involves the contention that all air volumes entenng a specific
internal area of a structure will instantly and perfectly mix with all other air volumes that
occupy that area. Its use permits one to ignore details of complex interactions between air
streams of differing flowrate and orientation, and force attention upon the overall average
contamination level expected
Earlier literature dealing with the dilution of airborne contaminants within an industrial
setting typically suggested that actual contaminant levels at specific sites may be greater than
the overall average by a factor in the range of 3 to 10 More recent work, however, has
shown that the "mixing factor" is more often on the order of 2, and sometimes considerably
less than 1 in typical rooms This latter finding is completely logical and recognizes the law
of mass conservation and its application to the scenarios of interest. In consequence, the
assumption of perfect mixing has the following ramifications:
• It ignores the potential adverse effects of "hotspots" within a given space,
and
• It ignores the potential benefits to the analysis of locations that are less
contaminated than average conditions
The assumption that the entire internal volume of a structure can be treated as a single
compartment is essentially an extrapolation of the perfect mixing assumption from the
boundaries of any specific room to the outer boundary of all building areas served by the
same ventilation system. The advantages and disadvantages of this action are analogous to
those itemized for perfect mixing
As formulated, the model does not address the possibility that the contaminant may
become deposited on exposed surface areas or somehow decay. In this respect, it may
provide conservative results for reactive or particulate contaminants but underestimate
exposures resulting from radioactive particles that have settled on internal surfaces (and
which may pose threats unrelated only to inhalation of toxic substances) The assumption of
a perfect air balance simply contends that the total volumetric rate of air entenng the
structure of interest is exactly equal to the rate of air leaving
The assumption that the external contaminant concentration is a constant is most
significant for short duration cloud passage, but can be mitigated by use of an average rather
than peak value for Q^ The assumption has little impact in the case of relatively long
duration plumes of contaminants evolved at a fairly constant emission rate Consideration of
time dependent source terms and resulting transient contamination levels within a structure
requires a substantially more sophisticated computational procedure, albeit one that would be
based on the same fundamental principles of the model presented herein.
C-12
-------
APPENDIX D
CHEMICAL COMPATIBILITY CHART
Introduction
Most information sources on the reactivity hazards of hazardous materials list and
describe the specific consequences of combining specific chemicals. The best and most
complete of these are cross-referenced compilations of reports of past mishaps which
occurred during experimentation or use of specific chemicals or substances, as reported in
the general literature. Two particularly well researched and readily available compilations
were referenced in Chapter 7.
Since there are literally tens of thousands of known chemicals in commerce, and since
the consequences of only a small fraction of the possible combinations of these materials
have been reported upon in the general literature, none of these information sources can
claim (and none do) that combinations of unlisted materials will not produce a hazardous
reaction. Thus, although these information sources provide valuable guideposts for evaluat-
ing potential chemical compatibility hazards, they are inherently limited in scope, and cannot
always be relied upon to provide the user with desired and/or necessary information.
To help fill gaps in available information and data with a reasonable amount of effort, a
relatively little known project was undertaken by the California Department of Health
Services some years ago to develop a single chemical compatibility chart which would
provide its user with a general indication of the typical effects of mixing a substance from
one chemical family with a substance from another family By concentrating on the most
common families of chemicals (i.e., substances with generally similar molecular structure),
and by studying the effects of combining the most reactive chemicals in each family,
researchers were able to produce a chart that provides an excellent tool to screen lists of
chemicals (as might be found at a facility or in a transportation vehicle) for those which may
pose an unusual and/or dangerous threat when inadvertently combined (as might occur
during some sort of process upset, transportation accident, or mismanaged material transfer
operation). Although a later bnef and apparently unpublished paper provides additional
information, the primary reference for the work being cited is:
• Hatayama, H.K, Chen, J J, de Vera, E.R., Stephens, R, and Storm, D L., A
Method of Determining the Compatibility of Hazardous Wastes, EPA Report
No. EPA-600/2-80-076, Municipal Environmental Research Laboratory, U.S
Environmental Protection Agency, Cincinnati, Ohio, Apnl 1980. Available
as publication PB80-221005 form the National Technical Information Ser-
vice, Springfield, Virginia 22161.
D-l
-------
This appendix describes and presents the chart discussed above for use by emergency
planning personnel in screening chemical substances in close proximity for potential
incompatibilities. Note that the chart was developed by investigation of reactions between
relatively pure chemical substances and therefore is not at all restricted to use for substances
only considered to be hazardous wastes Indeed, the chart is likely to be more reliable for
combinations of relatively pure chemicals than for multicomponent waste streams.
A Word of Caution
Before introducing the chart and describing how it is used, it is well to first present
somewhat modified and expanded versions of the warnings that its authors wished to provide
all potential users, these being:
The chart is intended to provide an indication of some of the hazards that can be
expected upon mixing of chemical substances Because of the differing activities
of the thousands of compounds that may be encountered, it is not possible to
make any chart definitive and all inclusive It cannot be assumed that members
of chemical families not listed on the chart will be compatible with each other or
-with listed families. Although blanks on the chart generally indicate that there is
no hazardous incompatibility expected between the families being considered, it
cannot be guaranteed that this will always be the case Defatted instructions as
to the hazards involved in handling and/or disposing of any given substance
should be obtained from the originator of the material or other expert source of
information.
The potential reaction consequences predicted by the chart are based on pure
chemical reactions only at ambient temperature and pressure Concentration,
synergistic, and antagonistic effects have been assumed not to influence the
reactions. The reactions have not yet been validated on actual waste materials
containing individual chemicals.
To the above caution must be added the observation that the chart is solely applicable
to the combination of two materials from different families. The addition of one or more
other materials to a mixture may (or may not) produce substantially different hazards. The
chart should not be used in any attempt to identify materials that may be self-reactive (i.e,
capable of runaway exothermic polymerization, runaway exothermic decomposition, explo-
sion, or other hazardous activity upon simple heating or exposure to air).
D-2
-------
Introduction to Table D.I
Although the chemical compatibility chart is presented as Figure D 1 and directly
follows this introductory text for convenience in future use, it is first necessary to discuss the
concept of Reactivity Group Numbers (RGNs) and to introduce Table D.I.
The compatibility chart addresses 41 classes or families of chemical substances, each
of which has been assigned a specific Reactivity Group Number that appears on the left and
bottom margins of the chart. Each of the first 34 classes or families are listed alphabetically
and are consecutively numbered one through 34. Members of each of these classes or
families generally share structural similarities with other members of their class or family at
the molecular level The last seven reactivity groups are numbered 101 to 107 and are
generic classes based on similarities in general behavior vis-a-vis chemical reactivity rather
than similarities in chemical structure or properties.
Since the first 34 classes or families are differentiated on the basis of the presence of a
particular type of molecular subgroup within the overall chemical molecule, and since the
other seven classes are relatively generic in nature, it is important to realize that any given
specific chemical can belong to more than one class or family. Indeed, some specific
materials may belong to several, exhibiting one or more of the reactive properties of each.
To assist readers in determining the appropriate Reactivity Group Numbers (RGNs) to
be assigned to any given material, Table D.I lists a large number of common chemical
substances together with their respective RGNs. The list of chemicals (including tradename
materials denoted with asterisks) in the table was compiled from several sources, with lists of
Hazardous Wastes and Hazardous Materials, Extremely Hazardous Wastes, and Extremely
Hazardous Materials denoted in California's Industrial Waste Law of 1972 (Ref. 44) serving
as a starting point. Other sources of information used to compile the list included references
1, 7, 10, 12, 13, 14, 32, 52, and 77 cited at the end of this appendix. The remaining
references were used to develop the chart and are presented as a resource for those who wish
to further investigate the reactivity of various substances.
Introduction and Use of the Chemical Compatibility Chart
As noted above, the actual chemical compatibility chart is presented in Figure D.I that
directly follows this introductory text The first column of the chart lists the RGNs and is
followed by a second column providing the respective names of the reactivity groups (i.e.,
the various classes and families of chemical substances) The RGNs are repeated on both the
top and bottom of each page of the chart (which could not be placed on one page while
maintaining readability) to facilitate finding of the specific square containing either a blank
or the Reactivity Codes (RCs) for specific combinations of classes or families
D-3
-------
Reactivity Codes (RCs) are abbreviations for the various hazardous phenomena
expected to occur when two materials are combined and are defined on the second page of
the chart They include letters or combinations of letters such as:
• H to indicate heat may be evolved.
• F to indicate potential for fire.
• £ to indicate potential for an explosion.
• P to indicate potential for violent polymerization.
• GT to indicate potential evolution of toxic gases.
• GF to indicate potential evolution of flammable gases.
• G to indicate potential evolution of innocuous gases.
• S to indicate solubilization of toxic substances.
• U to indicate an unknown but potentially hazardous reaction.
Where more than one code appears in a square, the one on the left (and generally the
highest) indicates the primary consequence of the combination, the next to the right indicates
the potential secondary effect, and so forth. The appearance of both GT and GF do not
necessarily mean that two different gases will evolve, since any of a number of gases may be
both flammable and toxic. (Note: As a precaution, it is well to treat the presence of a G in a
square with a grain of salt Until the gas is identified, treat it at least as being toxic. The
chart is good, but not necessarily perfect.)
Follow the instructions presented below to find the Reactivity Codes (RCs) for the
desired combination of two chemicals. Assume for the moment that each chemical has only
one RGN, but realize that the procedure may have to be repeated several times for substances
with multiple RGNs, once for each valid combination of RGNs.
Step 1: Using a knowledge of chemistry or the list of chemicals in Table D.I,
determine the Reactivity Group Number (RGN) appropriate for each of the
two chemical substances of concern. If a specific chemical is not found in
the table, look for one with similar characteristics and molecular groups
(usually indicated ~ but not always — by similar name fragments)
Step 2: For the first chemical, find its RGN on the first column of the chart
Step 3: Find the RGN of the second chemical on the bottom row of the chart.
Step 4: Find the intersecting square for the two RGNs. For example, the intersec-
tion of RGN 101 with RGN 2 has reaction codes of H plus F plus GT in the
correct square. The intersection of RGN 21 and RGN 20 has reaction codes
of GF and H in the correct square.
D-4
-------
Step 5: Note the Reaction Codes (RCs) m the square and refer to the chart legend
for their definition.
Step 6: If no RCs are in the square (i.e., the square is blank), the combination is
generally considered to be safe, keeping in mind the cautions presented
earlier.
D-5
-------
Re*ajyky
OroopNo.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
101
10Z
103
104
105
106
107
REACTIVITY GROUP NAME
AcJdiMineralNon-cxldlzmi
Add! Mineral Oxidizing
Add* Organic
Alcohol* and Olycofc
Aldehyde*
Amide*
talkie*. Aliphatic tod Aromatic
\xo Compound* t Diazo Compound!, and Hydrazlnc*
2atbamate*
Cauitica
Cyanide*
DJthiocaramate*
Bust*
Ethen
Fluoride*, Inorganic
Hydrocarbons, Aromatic
Halogeoatcd Organic*
Iiocyanete*
Ketooe*
Mercaptani «nd Other Organic Sulfide*
M«Ub, Alkali md Alkaline Earth, Elemental
Metali, Other Elemental & Alloyf a* Powden, V«pon,
orSpooge*
Metib , Other Element*! 4 Alloy» u Sheeti. Rod, Drop*,
Moldk?g«ett. ^ ^^
MeUk tod Metal Compounds, Toxic
Nitride*
Nitrite
Nttn Compound*, Orgmlc
Hydncuboct, Allphxde, Ununinted
Hydrocsrbou, Aliphatic, Saturated
Peroxide* tad Hydroperoxida, Organic
Pheooli and Crttok
Oixasopboiphate* , Photpbothioatei, Phoiphodidiloate*
SulHde*, Inorganic
Epoxidc*
Combuitible and Flammable Material!, Miscellaneoui
Exploilve*
Polymeriable Compound!
Oxidizing Ageoti, Strong
Reducing Agenti, Strong
Water and Mixture* Containing Water
Water Reactive Subftance*
1
H
Hp
H
H
«0
H0
H
°%
HOFp
H
H
OT
HQT
H0
H
01&F
°%F
0%F
°%F
S
<%
HogF
H
HO
H
HOT
01bF
HP
Ho
HE
PH
HOT
HOP
H
2
°H
Hp
HF
HOI
HOT
HOT
HOT
H
0%p
HOHF
Hp
Hp
OT
Hp
HR_
•OT
H K_
IGT
Hp
Hp
flT
OIHp
°\
OIHp
S
H B
FF
H p
FOF
H p_
•or
Hp
Hp
HE
Hp
HOT
HFQT
Hp
H u.
•far
HE
PH
H F
FOT
H
3
HP
HP
H
HO
H
01bF
H°&r
OT
HO
0!HF
OP
S
HOF
H
OT
Hp
HE
PH
HOT
HOP
4
HO
Hp
<%
OIHp
Hp
Hp
Hp
HOFp
FIGURE D.I
COMPATIBILITY CHART
S
H
H
H
OFOT
O!HF
O!HF
H
H
H0
H
U
Hp
Hp
6
<%
S
H p__
IOT
OFH
7
U
HOT
HP
OFH
S
HOT
Hp
H „
%T
HOF
8
°H
O
HO
H0
HO
"o
HO
Ho
0%
Hp_
%r
H n
F0
U
H a
FE
HO
U
E
Hp
HE
PH
HE
HO
o
9
HO
OIH
u
HO
Hp
•or
HE
H p
FOT
10
H
HOF
H p
P0
H
OFn
OFH
S
U
u
HE
HP
HE
PH
EXTREMELY REACTIVE
1
2
3
4
5
6
7
8
9
10
D-6
-------
Reactivity
Group No
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
101
1O2
103
104
105
IOC
107
11
H
HO
H
<%
°%
H%r
HP
*H
FOT
12
U
°H
<%
"•far
u
u
FOT
HOT
13
OIH
°H
HOFF
HOFp
OFH
Hp
HOFp
B
PH
HF
FB
OFH
DO NOT MK WITH ANY OffiNOCALCai WASTE MATERIAL
11
12
13 | 14
15
16
17
18
19
20
21
22
23
24
25
D-7
-------
Reactivity Code
Consequences
Example:
H Heat generation by chemical reaction
F Fire from extremely exothermic reaction, ignition of reaction products or
mixtures
G Innocuous and non-flammable gas generation such as Na, CO2, etc. causes
pressurization and rupture of closed containers
GT Toxic gas generation
GF Flammable gas generation
E Explosion due to extremely vigorous reaction or reactions produce enough
heat to detonate unstable reactants or reactions products
P Violet polymerization with extreme heat and sometimes toxic and
flammable gases
S Solubilization of toxic substances including metals
U May be hazardous but unknown
H
F Heat generation, fire, and toxic gas generation
GT
Reactivity
Group No.
26
27
24
29
90
31
32
33
34
101
102
103
104
105
106
107
26
H%T
H"OT
HGF
27
HE
HE
28
Hp
HE
29
HE
30
H
U
HQT
Hp
Hp
FOT
HE
PH
HO
HE
31
Hp
HE
PH
HE
°%
32
U
H n_
*OT
°K
33
Hp
HE
PH
H R_
•or
01OF
34
HE
HP
Fo
H
101
HE
HP
Fo
OFH
102
HE 103
HE H p^, 104
HE HPOP HPE ios
OFOT 106
EXTREMELY REACTIVE 107
26
27
28
29
30
31
32
33
34
101
102 103 104 105 106 107
D-8
-------
FIGURE D.2
RGN'S FOR SELECTED MATERIALS
Names
RGIS
Names
RGN
Abate*
Acenaphthene
Acetamide
Acetaldehyde
Acetic acid
Acetic anhydride
Acetone
Acetone cyanohydnn
Acetonitnle
Acetophenone
Acetoxy butane
Acetoxypentane
Acetyl acetone
Acetyl azide
Acetyl benzoyl peroxide
Acetyl bromide
Acetyl chloride
Acetylene
Acetyl nitrate
Acetyl peroxide
Acrolem
Acrylic acid
Acrylomtnle
Adipic acid
Adiponunle
Agallol
Agaloaretan
Aldicarb
Aldnn
Alkyl aluminum chloride
Alkyl resins
Allene
Allyl alcohol
Allyl bromide
Allyl chlonde
AUyl chlorocarbonate
Allyl chloroformate
Allyl tnchlorosilane
Aluminum
Aluminum aminoborohydnde
Aluminum borohydnde
Aluminum bromide
Aluminum carbide
32
16
6
5
3
107
19
4,26
26
19
13
13
19
102
30
17,107
17,107
28
27,102
30
5,103
3,103
26,103
3
26
24
24
9,20
17
107
101
28
4
17
17
13.17
13,17
107
22,23
107
105, 107
107
105
Aluminum chloride
Aluminum diethyl monochlonde
Aluminum fluonde
Aluminum hvdnde
Aluminum hvpophosphide
Aluminum phosphide
Aluminum tetiaazidoborate
Anunobenzene
Aminobutane
AminochJorotoluene
Armnodipheml
Aminoethane
Ammoethanol
Aminoethanolanune
AnunohexaiK
Aminomethane
Aminopentane
Aminophenol
Aminopropane
Ammo propiorutnle
Aminothiazole
Anunotoluene
Ammonia
Ammonium arsenate
Ammonium azide
Ammonium bifluonde
Ammonium chlorate
Ammonium dichromate
Ammonium fluonde
Ammonium hexamtrocobaltate
Ammonium hvdroxide
Ammonium hvpophosphide
Ammonium mohbdate
Ammonium nitrate
Ammonium rutndoosmate
Ammonium mtnte
Ammonium perchlorate
Ammonium penodate
Ammonium permanganate
Ammonium persulfate
Ammonium picrate
Ammonium sulfide
Ammonium letrachromale
107
105,107
15, 107
105
107
107
8
7
7
7,17
7
7
4.7
7
7
7
7
7,31
7
7,26
7.8
7
10
24
102
15
102, 104
24. 10?
15
24,10?
10
105
24
102
24, 104
102
104
102, 104
24, 102, 104
104
102
33.105
24.104
D-9
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
RGN
] L
Names
RGN
Ammonium tetnperoxychromate
Ammonium tnchromate
Amyl acetate
Amyl alcohol
Amyl chlonde
Arnylcyamde
Arnylimme
Amyleoe
Amylmercapun
Aniline
Ammert* V-101
Anisole
Anuole chlonde
Anthracene
Antimony
Antimony chlonde
Antimony fluonde
Antimony nitride
Antimony oxychlonde
Antimony oxide
Antimony pentachlonde
Antimony penUfluonde
Antimony penusulfide
Antimony perchlorete
Antimony poUMium urtrate
Antimony sulfate
Antimony tulfide
Antimony tnbrormde
Antimony tnchlonde
Antimony tnfluonde
Antimony tniodide
Antimony tnoxide
Antimony tnsulfale
Antimony tnsulfide
Antimony tnvmyl
Aqualin
Aqueous solutions & mixtures
Areun*
Aroclor*
Arsenic
Arsenic bromide
Arsenic chlonde
Arsenic disluflde
24, 102, 104
24.104
13
4
17
26
7
28
20
7
20
14
107
16
23,24
24,107
24,107
24,25
24
24
24
24
24, 33, 105
24,104
24
24
24, 33, 105
24,107
24,107
24,107
24.107
24
24
24.33
24,107
5.103
106
24
17
24
24,107
24,107
24, 33, 105
Arsenic iodide
Arsenic oxide
Arsenic penUselemde
Arsenic pentasulfide
Arsenic pentoxide
Arsenic sulfide
Arsenic tnbrormde
Arsenic tnchlonde
Arsenic tnfluonde
Arsenic tmodide
Arsenic tnsulfide
Arsine
Askarel
Asphalt
Azidocarbonyl guamdine
Azido-s-tnazole
Azinphos ethyl
Azindine
a,a-Azodusobutyronitnle
Azodnn*
Bakehte*
Hanoi
Banum
Banum azide
Banum bromate
Banum carbide
Banum chlorate
Banum chlonde
Banum chromate
Banum fluonde
Banum fluosilicate
Banum hydnde
Banum hydroxide
Banum hypophosphide
Banum icdate
Banum iodide
Banum monoxide
Banum nitrate
Banum oxide
Banum perchlorate
Banum permanganate
Banum peroxide
Banum phosphate
24,107
24
24
24,33
24
24,33,105
24,107
24,107
24
24,107
24, 33, 105
24,105
17
101
8,102
8
32
7,103
8,26
32
101
9
21, 24. 107
24,102
24,104
24, 105, 107
24,104
24
24,104
15,24
24
24,105
10,24
24,105
24,104
24
10, 24, 107
24,104
10, 24, 107
24,104
24,104
24,104
24
D-10
-------
FIGURE D.2 (Conk)
RGN'S FOR SELECTED MATERIALS
Names
RGN
J L
Names
RGN
Banum stearate
Banum sulfide
Banum sulfite
Bassa*
Bayer 25141
Baygon*
Benzadox
Benzal bromide
Benzal chlonde
Benzaldehyde
Benz-a-pyrene
Benzene
Benzene diazomum chlonde
Benzene phosphoras dichlonde
Benztdine
Benzoic add
Benzomtnle
Benzophenone
Benzoquinone
Benzotnazole
Benzotnbronude
Benzotnchlonde
Benzotnfluonde
Benzoyl chlonde
Benzoyl peroxide
Benzyl alcohol
Benzylamine
Benzyl benzene
Benzyl bromide
Benzyl chlonde
Benzyl chlorocarbonate
Benzyl chloroformate
Benzyl silane
Benzyl sodium
Beryllium
Beryllium copper alloy
Beryllium fluoride
Beryllium hydnde
Beryllium hydroxide
Beryllium oxide
Beryllium sulfide
Beryllium tetrahydroborate
Bidnn*
24
24, 33, 105, 107
24
9
32
9
6
17
17
5
16
16
8,102
107
7
3
26
19
19
8,102
17
17
17
107
30,102
4
7
16
17
17
17
17
105,107
105
24
24
15,24
24, 105, 107
10.24
24
33,105
24, 105, 107
32
Bismuth
Bismuth chromate
Bismuthic acid
Bismuth mtnde
Bismuth pentafluonde
Bismuth pentaoxide
Bismuth sulfide
Bismuth tnbromide
Bismuth trichloride
Bismuth tmodide
Bismuth tnoxide
Bismuth tnsulfide
Blada-fum*
Blue vitnol
Bomyl
Borane
Bordeaux arsemtes
Bone acid
Boron arsenotnbromide
Boron bromodiiodide
Boron dibromoiodide
Boron mtnde
Boron phosphide
Boron tnazide
Boron tnbromide
Boron tnchlonde
Boron tnfluonde
Boron tmodide
Boron tnsulfide
BPMC
Brass
Brormc acid
Bromine
Bromine aztde
Bromine cyanide
Bromine monofluonde
Bromine pentafluonde
Bromine tnfluonde
Bromoacetylene
Bromobenzoyl acetamlide
Bromobenzyl tnfluonde
Bromodiborane
Bromodiethylaluminum
22, 23, 24
24
24
24,25,102
24,107
24
24, 33, 105
24
24
24
24
24, 33, 105
32
24
32
24,107
24
1
24,105
24,107
24.107
24,25
24 107
24, l
24,107
24,107
24,107
24,107
24, 33, 105
9
23
2
104
102
11
104. 107
104,107
104,107
17
6,19
17
105
107
D-ll
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
RGN
Names
RGN
Bromodimethoxyaniline
Bromoform
Bromomethane
Bromophenol
Bromopropene
Bromopropyne
Bromotflane
Bromaoluene
Bromotrichloromethane
Bromotnfluomethane
Bromoxynil
Bronze
BunaN*
Bunker fuel oil
Butacarb
Butadiene
Butadiyne
Butanal
Butane
Bulanodiol
Buunethiol
Butanetnol tnmtnte
Buunol
Butanone
Butentl
Butene
Butene-2-one
Butyl acetate
n-Butyl acrylate
Butylamine
Butyl alcohol
t-Bulyl azidoformate
Butyl benzene
Butyl benzyl phthalate
Butyl cellusolve*
Butyl dichloroborane
Butyl ether
Butyl formate
Butyl fluonde
Butyl glycidyl ether
Butyl hydroperoxide
t Butyl hypochlontc
n-Butyl lithium
14
17
17
17,31
17
17
105
17
17
17
17,26.31
23
101
101
9
28,103
28
5
29
4
20
102
4
19
5
28
19
13
13,103
7
4
8
16
13
4
105
14
13
17
34
30
102,104
105, 107
Butyl mcrcaptan
Butyl peroxide
Butyl peroxyacetate
Butyl peroxybenzoate
Butyl peroxypivalate
t-Butyl peibenzoate
t-Butyl-3-phenyl oxaztrane
Butyl tnchloronlane
Butyranude
Butyraldehyde
Butync acid
Butyromtnle
Bux*
Cacodybc acid
Cadmium
Cadmium acetyhde
Cadmium amide
Cadmium aade
Cadmium bromide
Cadmium chlorate
Cadmium chlonde
Cadmium cyanide
Cadmium fluonde
Cadmium hexamine chlorate
Cadmium hexamine perchlorate
Cadmium iodide
Cadmium nitrate
Cadmium mtnde
Cadmium oxide
Cadmium phosphate
Cadmium sulfide
Cadmium tnhydrazine chlorate
Cadmium tnhydrazine perchlorate
Calcium
Calcium arsenate
Calcium arsenite
Calcium bromate
Calcium carbide
Calcium chlorate
Calcium chlorite
Calcium fluonde
Calcium hexammoraate
Calcium hydnde
20
30
30
30
30
30
34
107
6
5
3
26
9
24
23,24
24. 105, 107
24, 10, 107
24,102
24
24,104
24
11,24
15,24
24,102
24,102
24
24, 102, 104
24, 25, 102
24
24
24. 33, 105
24,102
24,104
24,102
24
24
104
105,107
104
104
15
105
105, 107
D-12
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
RGN
Names
RGN
Calcium hydroxide
Calcium hypochlonte
Calcium hypophosphide
Calcium lodate
Calcium-manganese-silicon alloy
Calcium nitrate
Calcium oxide
Calcium oxychlonde
Calcium peichromate
Calcium permanganate
Calcium peroxide
Calcium phosphide
Calcium sulfide
Camphor oil
Capnc acid
Caproic aad
Caprylic acid
Caprylyl peroxide
Carbacrol
Carbaryl
Carbetanude
Carbanolate
Carbofuran
Catbobc acid
Carbohc oil
Carbon, activated, spent
Carbon bisulfide
Carbon disulfidc
Carbon tetrachlonde
Carbon tetrafluonde
Carbon tctraiodide
Castnx
Catechol
Caustic potash
Caustic soda
CDEC
Cellulose
Cellulose nitrate
Cerium
Cenum hydride
Cerium tnsulfide
Cerous phosphide
Cesium
10
104
105
104
23
104
10,107
104
104
104
104
107
33,105
101
3
3
3
30
31
9
6
9
9
31
31
101
20
20
17
17
17
7
31
10
10
12
101
27,102
22
105
33,105
105
21
Cesium amide
Cesium azide
Cesium carbide
Cesium fluoride
Cesium hexahydroaluminate
Cesium hydride
Cesium phosphide
Cesium sulfide
Chloral hydrate
Chlordane
Chlorcstol
Chlorfenvinphos
Chloric acid
Chlorine
Chlorine azide
Chlorine dioxide
Chlorine fluoroxide
Chlorine monofluonde
Chlorine monoxide
Chlorine pentafluonde
Chlorine tnfluonde
Chlorine tnoxide
Chloroacetaldehyde
Chloroacetic acid
Chloroacetone
Chloroacetophenone
Chloroacetyl chloride
Chloroacetylene
Chloroacrylomtnle
Chloroazodin
Chlorobenzene
Chlorobenzotnazole
Chlorobenzoyl peroxide
Chlorobenzyhdene malonomtnle
Chlorobutyronitnle
Chloro chromic anhydnde
Chlorocreosol
Chlorodiborane
Chlorodiisobutyl aluminum
Chlorodimethylamine diborane
Chlorodinitrobenzene
Chloro duutrotoluene
Chlorodipropyl borane
107
102
105
15
105
105,107
107
33,105
5
17
17
32
2,104
104
102
102, 104, 107
102.104
104,107
104
104,107
104,107
102,104
5,17
3,17
17,19
17,19
107
102
17,26
8,17
17
8,17
17,30
17,26
17,26
24, 104, 107
17,31
105
105,107
105
17,27
17,27
105
D-13
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
Chloroethine
Chlorocthinol
Chloroethylerumine
Chloroform
Chlorohydnn
Chloromethane
Chloromethyl methyl ether
CWorom ethyl phenoxyaceuc acid
CUorotutrotnfline
Chloronitrobenzene
Chloropentane
Chlorophenol
Chlorophenyl isocyanate
Chloropicnn
CMoropropme
Chloropropene
Chloropropylene oxide
Giloroiilane
Chlorosulfomc icid
Chlorothion*
Chlorotoluene
Chlorotolindine
Chlorotnnitrobcnzenc
B-Chlorovinyldichloroarsine
Chlotpicnn
Chromic acid
Chromic uiyhdnde
Chromic chlonde
Chromic fluoride
Chromic oxide
Chromic sulfate
Chromium
Chromium lulfate
Chromic tulfide
Chromium tnchlonde
Chromium tnfluonde
Chromium tnoxide
Chromyl chlonde
Chrysene
CMME
Coalod
Coal tar
Cobalt
RGN
17
4,7
17
17
17
17
17
3,17
17,27
17,27
17
31
17, 18, 107
17,27,102
17
17
17.34
105
1
17,32
17
7,17
17, 27, 102
24
17,27,102
2,24,104
2,24,104
24
15,24
24
24
23,24
24
24, 33, 105
24
15,24
2,24,104
24, 104, 107
16
14,17
101
31
22, 23, 24
Names
RGN
Cobalt bromide
Cobalt chlonde
Cobalt nitrate
Cobaltous bromide
Cobaltous chlonde
Cobaltous mtnte
Coibaltous resinate
Cobaltous sulfate
Cobalt resinate
Cobalt sulfate
Collodion
Copper
Copper acetoarserute
Copper acetylide
Copper arsenate
Copper arserute
Copper chlonde
Copper chlorotetrazole
Copper cyanide
Copper nitrate
Copper mtnde
Copper sulfate
Copper sulfide
Compound 1836
Coroxon*
Coumafuryl
Coumatetralyl
Cresol
Cresol glydicyl ether
Cresote
Cnmidine
Cro ton aldehyde
Crotyl alcohol
Crotyl bromide
Crotyl chlonde
Cumene
Cumene hydroperoxide
Cupnc arsenate
Cupnc arserute
Cupnc chlonde
Cupnc cyanide
Cupnc nitrate
Cupnc sulfate
24
24
24.104
24
24
24,104
24
24
24
24
27
23.24
24
24. 102, 105, 107
24
24
24
24
11,24
24,104
24,25
24
24.33.105
17.32
32
19
19
31
34
31
7
5
4
17
17
16
30
24
24
24
11.24
24,104
24
D-14
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
T
RGN
Names
RGN
Cupnethylenediamine
Cyanoacetic acid
Cyanochloropentane
Cyanogen
Cyanogen bromide
Cyanophenphos
Cyanunc tnazide
Cycloheptane
Cyclohexane
Cyclohexanol
Cyclohexanone
Cyclohexanone peroxide
Cyclohexylamme
Cyclohexenyl tnchlorosilane
Cyclohexyl phenol
Cyclo'. sxyl tnchlorosilane
Cyclopentane
Cydopentanol
Cyclopentene
Cyclopropane
Cyclotnmethylene tnnitraamine
Cymene
Cyolan*
2.4-D
Dasamt*
DBCP
DCS
ODD
DDNP
DDT
DDVP
DEAC
Decaborane
Decahydronaphthalene
Decabn
Decane
Decanol
Decene
Decyl benzene
Delnav*
Demeton-s-methyl sulfoxid
Chacetone alcohol
Diacetyl
7,24
3,26
17,26
26
11
26.32
102
29
29
4
19
30
7
107
31
107
29
4
28
29
27,102
16
20,32
3,17
32
17
17
17
8,27,102
17
17,32
105,107
107
29
29
29
4
28
16
32
32
4,19
19
Diacetylene
Diamine
Diaminobenzene
Diammohexane
Diaadoethane
Diazinon*
Diazoduutrophenol
Dibenzoyl peroxide
Diborane
Diboron hexahydnde
Dibutyl ether
Dibutyl phthalate
3,5-Dibromo-4-hydroxybenzonitnle
Dibromochloropropanc
Dibromoe thane
Dichloroacetone
Dichloroamine
Dichlotorobenzene
Dichlorobenadine
Dichlorodunethylsilane
Dichloroe thane
Dichloroethene
Dichloroether
DichloFoethylarsine
Ethyl dichlorosjane
Ethyl ether
Dichlormsocyanunc acid
Dichloromethane
Dichlorophene
Dichlorophenol
Dichlorophenoxyacetic acid
Dichloropropane
Dichloropropanol
Dichloropropene
Dichloropropylene
Dichloro-s-tnazine-2,4,5-tnone
Dichlorovos
Dicumyl peroxide
Dicyclopentadiene
Eheldnn
Diethanolamine
Diethyl aluminum chlonde
3iethylanune
28
8,105
7
7
8,102
32
27,102
30,102
105,107
105,107
14
13
17.26.31
17
17
17,19
104
17
7,17
107
17
17
14,17
24,107
107
14,17
104
17
17
17,31
3,17
17
4,17
17
17
104
17,32
30
28
17
4,7
105,107
7
D-15
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
RGN
] L
Names
RGN
Diethyl benzene
Diethyl chlotovinyl phosphate
Diethyl dichlorosilane
Dieihylene dioxide
Diethylene glycol dmitrale
Diethylene glycol monobutyl ether acetate
Dieihylene tnamme
Diethyl ether
Diethyl Icctone
Diethyltoluamide
Diethyl zinc
DieieloU
Difluorophoiphonc acid
Diglyadyl ether
Diisobutylene
Diisobutyl ketone
Diisopropanolamine
Dntopropylbenzene hydroperoxide
Diitopropyl beryllium
Diisopropyl ether
Duiopropyl peroxydicarbonate
Dimecron*
Dunefox
Dimethyl acetylene
Dimethyl amine
Dunethylammo azobenzene
Dimethyl arsenic acid
Dimethyibenzyl hydroperoxide
Dimethyl butane
Dimethyl butyne
Dimethyl dichlorosilane
Dimethyldilhiophosphonc acid
Dimethyl ether
Dimethyl formal
Dimethyl formamide
Dimethylhexane dihydroperoxide
Dimethyl hydiazine
Dimethyl ketone
Dimethyl magnesium
Dimethylmtrobenzene
Dimethylnitrosoamine
Dimethyl sulfide
Dimeton
16
17.32
107
14
27.102
13
7
14
19
6
24, 105, 107
101
1
34
28
19
4.17
30
24, 104, 107
14
30
32
6,32
28
7
7,8
24
30
29
28
107
32
14
19
6
30
8
19
105,107
27
7,27
20
09
Dinitrobenzene
Dimtrochlorobenzene
2,4-Dimtro-6-sec-butyl phenol
Dimtrocresol
Duutiophenol
Din ''rophcnyl hydrazine
Dinitrotoluene
Dinoseb
Dioxacarb
Dioxane
Dioxathion
Dipentaerythntol hexamtrate
Dipentene
Diphenamide
Diphenyl
Diphenyl acetylene
Diphenylamine
Diphenylamine chloroarsine
Diphenyl ethane
Diphenyl ethylene
Diphenyl methane
Diphenylmethane diisocyanate
Diphenyl oxide
Dipicryl amine
Dipropyl amine
Disulfoton
Disulfunc acid
Disulfur dimtnde
Disulfuiyl chlonde
Disyston*
Dithane* M-45
Dithione*
DNOC
Dodecene
Dodecyl benzene
Dodecyl tnchlorosilane
Dowco-139*
Dowicide I
Dowtherm
Durene
Dyfonate*
Dynes Thinner
'•'eetolSO
27
17,27
27,31
27,31
27,31
8,27
27
27.31
9
14
32
27,102
28
6
16
16
7
7,24
16
16
16
18,107
14
7,27,102
7
32
1
25,102
107
32
12
32
27,31
28
16
107
9
31
16
16
32
101
27,31
D-16
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
RGN
Names
Endolsulfan
Endothall
Endothion
Endnn
EPN
Epichlorohydnn
Epoxybutane
Epoxybutene
Epoxyethane
Epoxyethylbenzene
Bis(2-3-Epoxypropyl) ether
Ethane
Ethanethiol
Ethanol
Ethion*
Ethoxyethanol
Ethyl acetate
Ethyl acetylene
Ethylacrylate
Ethyl alcohol
Ethylamine
Ethyl benzene
Ethyl butanoate
Ethyl butyrate
Ethyl chloride
Ethyl chloroformate
Ethyl dichloroarsine
Ethyl dichlorosilane
Ethyl ether
Ethylene
Ethylene chromic oxide
ethylene chlorohydnn
Ethylene cyanohydnn
Ethylene diamine
Ethylene dibromide
Ethylene dichlonde
Ethylene glycol
Ethylene glycol duutrate
Ethylene glycolmonomethyl ether
Ethyleneumne
Ethylene oxide
Ethyl formate
2-Eihylhexyl acrylate
17,20
3
32
17
32
17.34
34
34
34,103
34
34
29
20
4
32
4.14
13
28
13.103
4
7
16
13
13
17
13,17
24,107
107
14
28
24,104
4,17
4.26
7
17
17
4
27.102
4, 14. 17
7,103
34.103
13
13.103
RGN
Ethyl mercaptan
Ethyl nitrate
Ethyl rutnte
Ethyl pnpionate
Ethyl tnchlorosilane
Exothion
Eugenol
Fensulfothion
Ferbam
Feme arsenate
Feme sulfide
Ferrous arsenate
Ferrous sulfide
Fluoranthrene
Fluorene
Fluorine
Fluorine azide
Fluorine monoxide
Fluoroacetanihde
Fluoroacetic acid
Fluorobonc aad
Fluorosulfomc acid
Fluosulfomc aad
Fluosihac acid
Fonofos*
Formaldehyde
Form amide
Formetanate hydrochlonde
Formic acid
Fostion*
Freon*
Fumanc acid
Fumann
Fumazone*
Furadan*
Furan
Furfural
Furfuraii
Gas oil, cracked
Gasoline
Germanium sulfide
Glutaraldehyde
Glycerin
20
27,102
27,102
13
107
32
31
32
12
24
33
24
33,105
16
16
104,107
102
104,107
6,17
3
1,15
1,107
1,107
1,15
32
5
6
6
3
32
17
3
19
17
9
14
5
14
101
101
33,105
5
4
D-17
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
Glycidol
G!>col diacetate
Gljcoldinitntc
Glycol ether
Glycolic acid
Glycol monohc J.c tnmtrate
Glycol omtnle
Gold acelyhde
Gold cyanate
3old fulnunaie
Gold sulflde
Grease
3uaiacol
Guanyl nitrosar—noguanylidene hydrarine
Guarudine nitrate
Gun cotton
Guthion*
Hafnium
Hanane*
Kcmimellitene
Hcpuchlor
Heptane
Hcpunal
Hcptanol
Hepunone
Hcptene
Hcxaborane
Hexachlorobcnzsne
Hexadecyl tncho-osilane
Hexaethyl tetrap^osphate
Hexafluorophoiphonc aad
Hexahydnde diborane
Hexamethyl benzene
HexamcthylcncdiaTune
Hexamethylenc : raamine
Hcxanal
Hcxanitrodjphcr,\'amine
Hexanol
Ilcxanoic acid
Hcxcne
Hex>lamine
Hex>l tnchlorosi-ine
Hcxyne
RGN
34
13
27,102
14
3
27,102
26
105, 107
102
102
33, 105
101
31
8,102
27,104
27,102
32
22
6,32
16
17
29
5
4
19
28
105
17
107
32
1,15
105, 107
16
7
7
5
7,27,102
4
3
28
7
107
28
Names
HMX
•lopcide*
•lydrated lime
•lydrazine
Hydrazine azide
lydrazoic acid
fydnodic acid
rtydrobromic acid
Hydrochloric acid
Hydrocyanic acid
Hydrofluoric acid
Hydrogen azide
Hydrogen bromide
Hydrogen cyanide
Hydrogen fluoride
Hydrogen iodide
Hydrogen peroxide
Hydrogen phosphide
Hydrogen selenide
Hydrogen sulfide
Hydroqumone
Hydroxyacetophenone
Hydroxydibromobenzoic acid
Hydroxydiphenol
Hydroxyhydroquinone
Hydroxyacetophenone
Hydroxyisobutyronitnle
Hydroxyl amine
Hydroxypropioratirle
Hypochlorous acid
Indcne
Indium
Ineneen
Iodine monochlonde
Iodine pentoxide
Iron
Iron arsenate
Isobutane
Isobulanol
Isobutyl acetate
Isobutyl acrylate
Isobutylene
Isodecyl acrylate
RGN
102
9
10
8,105
8,102
102
1
1,107
1
1,11
1.15
102
1,107
1,11
1,15
1
104
105
24, 105
33. 105
31
19,31
3,17
31
31
19,31
4,26
105
4,26
2
16
22, 23, 24
17
107
104
23
24
29
4
13
13, 103
28
13
D-l*
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
Isodurene
Isoeugenol
Isohexane
Isooctane
Isooctene
Isopentane
Isophorone
Isoprcne
Isopropanol
Isopropyl acetate
Isopropyl acetylene
Isopropylarrane
Isopropyl benzene
Isopropyl chlonde
Isopropyl ether
Isopropyl mercaptan
N-Isopropylmethylcarbamate
a-Isopropyl methylphosphoiyl fluonde
Isopropyl percarbonate
Isotacnc propylene
MOO
Jet oil
Kerosene
Lacquer thinner
Landnn*
Lannate*
Lauroyl peroxide
Lead
Lead acetate
Lead arsenate
Lead arserute
Leadazide
Lead carbonate
Lead chlorite
-ead cyanide
jead dimtroresorcinate
Lead mononitroresorcmate
jead nitrate
Lead orthoarsenate
Lead oxide
Lead styphnate
Lead sulfide
jead tnmtroresorcinate
RGN
16
31
29
29
28
29
19
28,103
4
13
28
7
16
17
14
20
9
17,32
30
101
101
101
101
101
9
9,20
30
23,24
24
24
24
24,102
24
24,104
11,24
24, 27, 102
24. 27, 102
24,104
24
24
24, 27, 102
24, 33, 104
24, 27. 102
Names
Lewisite
Lime nitrate
Lindane
Lithium
Lithium aluminum hydride
Lithium amide
Lithium ferrosihcon
Lithium hydride
Lithium hydroxide
Lithium hypochlonte
Lithium mtnde
Lithium peroxide
Lithium silicon
Lithium sulfide
London purple
Lye
Magnesium
Magnesium arsenate
Magnesium arsemte
Magnesium chlorate
Magnesium fluoride
Magnesium nitrate
Magnesium perchlorate
Magnesium peroxide
Vlagnesium sulfide
Malathion
Maleic add
Maloruc nitrite
Maneb
Manganese
Manganese acetate
Manganese arsenate
Manganese bromide
Manganese chlonde
Manganese methylcyclopentadienyltncarbonyl
Manganese nitrate
Manganese sulfide
Manganous arsenate
Manganous bromide
Manganous chlonde
Manganous nitrate
Manmtol hexamtrate
Matacil*
RGN |
24
104
17
21,107
105,107
10,107
107
105, 107
10
104
25
104,107
107
33,105
24
10
21,22
24
24
104
15
104
104
104
33.105
32
3
3,26
12
22, 23, 24
24
24
24
24
24
24,104
24, 33, 105
24
24
24
104
27,102
9
D-19
-------
FIGURE D.2
RON'S FOR SELECTED MATERIALS
Names
uliyer "s reagent
ri edinoterb acetate
Meobtl
rlercaptobenzothiazole
ifercatoethanol
vlercarbam
rfercunc acetate
rfereunc ammonium chlonde
Mercuric benzoate
tfcrcunc bromide
ritrcunc chlonde
.Icicunc cyanide
riercunc dioxysulfate
vlercunc iodide
vlercunc nitrate
Mercuric oleate
Mercuric oxide
Mercuric oxycyanide
Mercunc potassium iodide
Mercuric salicyiate
Mercunc subiulfate
Mercunc sulfate
Mercunc sulfide
Mercunc thiocyanate
Mercunc thiocyamde
Mercuro!
Mercurous bromide
Mercurous gluconate
Mercurous iodide
Mercurous nitrate
Mercurous oxide
Mercurous sulfate
Mercury
Mercury (vapor)
Mercury acetate
Mercury ammonium chlonde
Mercury benzoate
Mercury bisulfate
Mercury chlonde
Mercury cyanide
Mercury fulminate
Mercury iodide
Mercury nitrate
RGN
24
13,27
9
8,20
4,20
32
24
24
24
24
24
11,24
24
24
24,104
24
24
11.24,102
24
24
24
24
24. 33, 105
24
24
24
24
24
24
24,104
24
24
24
22.24
24
24
24
24
24
11,24
24.102
24
24,104
Names
Mercury nucleate
Mercury oleate
Mercury sulfate
Mesitylene
Mesityl oxide
Mesurol*
Metasystox-R
Metham
Methanal
Methane
Methanethiol
Methanoic acid
Methane!
Methomyl
Methoxyethylmercunc chlonde
Methyl acetate
Methyl acetone
Methyl acetylene
Methyl acrylate
Methyl alcohol
Methyl aluminum sesquibronude
Methyl aluminum sesquichlonde
Methylamine
Methyl amyl acetate
N-Methyl aniline
Methyl azindine
Methyl benzene
Methyl bromide
Methyl butadiene
Methyl butane
Methyl butene
Methyl butyl ether
Methyl t-butyl ketone
Methyl butyne
Methyl butyrate
Methyl chlonde
Methyl chlorocarbonate
Methyl chloroform
Methyl chloroformate
Methyl chloromethyl ether
Methyl cyanide
Methyl cyclohexane
Methyl dichloroaisine
RGN
24
24
24
16
19
9
32
12
5
29
20
3
4
9,20
24
13
101
28
13,103
4
105. 107
105,107
7
13
7
7
16
17
28,103
29
28
14
19
28
13
17
13,17
17
13,17
14,17
26
29
24
D-20
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
RGN
Names
Methyl dichlorosilane
Methylene chlonde
Methylene diisocyanate
4,4-Methylene bis(2-chloroaniline)
Methyl ethyl chlonde
Methyl ethyl ether
Methyl ethyl ketone
Methyl ethyl ketone peroxide
Methyl ethyl pyndine
Methyl formate
Methyl hydrazme
Methyl iodide
Methyl uobutyl ketone
Methyl isocyanate
Methyl isopropenyl ketone
Methyl magnesium bromide
Methyl magnesium chlonde
Methyl magnesium iodide
Methyl mercaptan
Methyl methacrylate
Methyl napthalene
Methyl parathion
Methyl pcntanoate
Methyl propionate
Methyl n-propyl ketone
Methyl styrene
Methyl sulfide
Methyl tnchlorosilane
Methyl valerate
Methyl vinyl ketone
Methyl yellow
Mevinphos
Mexacarbate
Mineral spints
Mintacol*
Mipcin*
Mobam*
Mocap*
Molybdenum
Molybdenum anhydride
Molybdenum sulfide
Molybdenum tnoxide
Molybdic acid
107
17
18,107
7,17
17
14
19
30
7
13
8
17
19
18,107
19
105.107
105,107
105, 107
20
13,103
16
32
13
13
19
28,103
20
107
13
19
7,8
32
9
101
32
9
9
32
22,23,24
24
24,33,105
24
24
RGN
Monochloroacetone
MonochloroaceUc acid
Monocroptophos
Monoethanol amine
Monofluorophosphonc acid
Monoisopropanolanune
Monomethyl hydrazine
Motphohne
Municipal solid waste
Munatic acid
Nabara
Nack
Nak
Naptha
Naphthalene
Napthol
Napthylanune
Naphthyl mercaptan
Naphtite
Nemagon*
Neohexane
4-NBP
Niacide*
Nialate
Nickel
Nickel acetate
Nickel antimomde
Nickel arsenate
Nickel arsenite
Nickel carbonyl
Nickel chlonde
Nickel cyanide
Nickel nitrate
Nickelous arsenate
Nickelous arsenite
Nickelous chlonde
Nickelous nitrate
Nickel selemde
Nickel subsulfide
Nickel sulfate
Nickel tetracarbonyl
tfitrambne
Mitnc acid
17,19
3,17
32
4.7
1
4,7
8
7
101
1
12
21,107
21,107
101
16
31
7
20
27,102
17
29
27
12
32
22,24
24
24,107
24
24
24
24
11,24
24.104
24
24
24
24,104
24
24, 33, 105
24
24
7,27
2
D-21
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names RGN
silrotniline
Nitrobenzene
fttrobcnzol
ftuobiphenyl
tarocalaty
ftiroceUulote
filrochlotobenzene
Nitrogen dioxide
ifiuominnite
n'Urogen mujttrd
Nitrogen tetroxide
iitroglycenn
•< itrohy drochlonc tad
v'itrophaiol
"Jiiroproptne
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
RGN
Phenylaniline
Phenylbenzene
Phenylbutane
Phenylchloromethyl ketone
Phenyl dichloroarsine
Phenylene diaimne
Phenylethane
Phenyl hydrazme hydrochlonde
o-Phenyl phenol
Phenyl tnchlorosilane
Phenyl valerylmtnle
Phenylpropane
Phloroglucinol
Phorate
Phosdnn*
Phosphanudon
Phosphine
Phospholan
Phosphoraum iodide
Phosphoric acid
Phosphonc anhydnde
Phosphonc sulfide
Phosphorus (Amorphous red)
Phosphorus (White- Yellow)
Phosphorus heptasulfide
Phosphorus oxybromide
Phosphorus oxychlonde
Phosphorus pentachlonde
Phosphorus pentasulfide
Phosphorus pentoxide
'hosphorus sesquisuffide
Phosphorus tnbromide
'hosphorus tnchlonde
'hosphorus tnsulfide
'hosphoryl bromide
'hosphoryl chlonde
Phthalic acid
'icramide
'icnc acid
'icndine
'icryl chlonde
Pipendme
'inrmcarb
7
16
16
17,19
24
7
16
8
31
107
26
16
31
32
32
32
105
20,32
105.107
1
107
33, 105, 107
105,107
105
33,105
104,107
104, 107
107
33, 105, 107
107
33, 105, 107
107
107
33, 105, 107
104, 107
104,107
3
7,27,102
27,31,102
7
17, 27, 102
7
9
Names
Polyglycol ether
Polyanude resin
Polybronunated biphenyl
Polybutene
PolycMonnated biphenyls
Polychlormated tnphenyls
Polethylene
Polyester resin
Polymenc oil
Polyphenyl polymethylisocyanate
Polypropylene
Polyram combi*
Polysulfide polymer
Polystyrene
Polyurethane
Polyvjnyl acetate
Polyvinyl chlonde
Polyvmyl nitrate
Potasan
Potassium
Potassium acid fluonde
Potassium alununate
Potassium arsenate
Potassium arsemte
Potassium bifluonde
Potassium bichromate
Potassium bromate
Potassium butoxide
Potassium cyanide
Potassium dichloroisocyanurate
Potassium dichromate
Potassium dinitrobenzfuroxan
Potassium fluonde
Potassium hydride
Potassium hydroxide
Potassium nitrate
Potassium nitnde
Potassium mtnte
Potassium oxide
Potassium perchlorate
Potassium permanganate
'olassium peroxide
'otassium sulfide
RGN
14
101
17
28
17
17
101
101
101
18,107
28, 101
12
20. 101
101
101
101
101
27,102
32
21,107
15
10
24
24
15
24,104
104
10
11
104
24,104
27,102
15
105, 107
10
102,104
25
104
107
104
24,104
104, 107
33,105
D-23
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
RGN
'romecarb
topanal
'roptne
*rop*nelhiol
"roptnoic acid
'roptnol
•ropirgyl bromide
'roptrgyl chloride
:-Propen-l-ol '
'roppiolactone ^
'ropion aldehyde
'ropion amide
'ropionic acid
'ropionitnle >
Propylaceute N
Propyl alcohol
'ropyUmine
?ropyl benzene
Propylene dichlonde
Propyleneglycol
Propylene glycolmonomelhyl ether
Propylene oxide
Propylenammc
ftsgylahcr
Pr^ylfeimiie
Propyl mcrcipun
Propyl Tnchlorosilme
Prothotte
Pteudocumene
Pyndine
Pyrogallol
Pyrosulfuryl chlonde
Pyroxylin
Quinone
Raney nickel
RDX
Refuse
Resins
Resorcinol
Rubidium
Salicylated mercury
S nilgai in
Saltpeter
9
5
29
20
3
4
17
17
4
13
5
6
3
26
13
4
7
16
17
4
4,14
34.103
7
14
13
20
107
32
16
7
31
107
27
19
22
27,102
101
101
31
21
24
31
102, 104
Names
Schradan
Selemous acid
Selenium
Selenium diethyldithiocarbamate
Selenium fluoride
Seiaiousacid
SJicochloroform
Silicon tetrachonde
Silicon tetrafluonde
Silver acetylide
Silver azide
Silver cyanide
Silver nitrate
Silver mtnde
Silver styphnate
Silver sulfide
Silver tetrazene
Silver trinitroresorcinate
Slaked lime
Smokeless powder
Sodamide
Soda niter
Sodium
Sodium acid fluoride
Sodium aluminate
Sodium aluminum hydnde
Sodium amide
Sodium arsenate
Sodium arsemte
Sodium azide
Sodium bichromate
Sodium bifluonde
Sodium bromate
Sodium cacodylate
Sodium carbonate
Sodium carbonate peroxide
Sodium chlorate
Sodium chlorite
Sodium chromate
Sodium cyanide
Sodium dichloroisocyanurate
Sodium dichromale
Sodium dimethylarsenate
RGN
6,32
1,24
22,23,24
12,24
15.24
1,24
107
107
15,107
24, 102, 105, 107
24,102
11,24
24,104
24,25,102
24,27,102
24,33.105
24,102
24,27,102
10,107
102
10,107
104
21, 105, 107
15
10,105
105,107
10,107
24
24
102
24,104
15
104
24
10
104
104
104
24
11
104
24,104
24
D-24
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
RGN
Sodium fluonde
Sodium hydride
Sodium hydroxide
Sodium hypochlonte
Sodium hyposulfite
Sodium methylate
Sodium methoxide
Sodium molybdate
Sodium monoxide
Sodium nitrate
Sodium rutnde
Sodium mtnte
Sodium oxide
Sodium pentachlorophenate
Sodium perchlorate
Sodium permanganate
Sodium peroxide
Sodium phenolsulfonate
Sodium picramate
Sodium polysulfide
Sodium potassium alloy
Sodium selenate
Sodium sulfide
Sodium thiosulfate
Stannic chlonde
Stannic sulfide
Starch nitrate
Solbene
Stoddard solvent
Strontium
Strontium arsenate
Strontium dioxide
Strontium monosulfide
Strontium nitrate
Strontium peroxide
Strontium tetnsulfide
Styphmc aad
Styrene
Succimc acid
Succimc acid peroxide
Sulfonyl chlonde
Sulfonylflounde
Sulfotepp
15
105, 107
10
10.104
105
10,107
10,107
24
10,107
104
25
104
10.107
31
104
24,104
104, 107
31
27,102
101
21,107
24
24, 33, 105
105
24,107
33,105
27,102
16
101
24
24
24.104
24, 33. 105
24.104
104
24, 33, 105
27,31.102
16, 28. 103
3"
30
107
107
32
Names
Sulfur chloride
Sulfur (elemental)
Sulfunc acid
Sulfunc anhydride
Sulfur monochlonde
Sulfur mustard
Sulfur oxychlonde
Sulfur pentafluonde
Sulfur tnoxide
Sulfuryl chlonde
Sulfuryl fluonde
Supracide*
Surecide*
SyntheUc rubber
TCDD
TEDP
TEL
TEPA
TEPP
THF
TMA
TML
TNB
TNT
Tall oil
Tallow
Tar
Tellurium hexafluonde
Tenuk*
Tetraborane
Tetrachlorodibenzo-p-dioxin
Tetrachloroe thane
Tetrachloroethylene
Tetrachloromethane
Tetrachlorophenol
retrachloropropyl ether
Tetradecene
Petraethyl dithionopyrophosphate
Tetraethyl lead
fetraelhyl pyrophosphate
^etrahydrofuran
'etramethylenediamine
etramethyl lead
RGN
107
101
2.107
104,107
107
20
107
15,107
104,107
107
107
32
32
101
14,17
32
24
6,32
32
14
7
24
27,102
27,102
101
101
101
15,24
9,20
105
14,17
17
17
17
17,31
14,17
28
32
24
32
14
7
24
D-25
-------
FIGURE D.2 (Cont.)
RON'S FOR SELECTED MATERIALS
Names
Tetiamethyl sucononitnle
Tetranitromethane
Tetraphenyl cthylcne
Teutphosphotus tniulfide
Tetnuelenium tetruutndc
Tetwul
Tetraiulfur tetramtnde
Tetnzene
Thallium
Thallium mtndc
Thallium sulfide
Thtllous lulfate
Thimet*
Thionyl chlonde
Thiocaibonyl chlonde
Thiodan*
Thionazin
Thionyl chlonde
Thiophosgene
Thiophosphoiyl chlonde
Thiram
Thonum
Tin tewchlonde
Titanic chlonde
TiUmum
Titanium seiquisulfide
Titanium sulfate
Titanium sulfide
Titanium tetnchlonde
TMA
TNB
TNT
Tblualdchyde
Toluene
Toluene dusocyanate
Toluic acid
Toluidmc
Toluol
Topade'
Tianid*
Tnamphos
Tnbromomethane
Tn n butylaluminum
RGN
26
27,102
16
33, 105, 107
24,25,102
20
25,102
8,102
24
24,25,102
24, 33, 105
24
32
107
107
17,20
32
107
107
107
12
22,23,24
24,107
24,107
22,23,24
24, 33, 105
24
24, 33, 105
24,107
7
27,102
27,102
5
16
18,107
3
7
16
6
9.26
6,32
17
107
Names
Tncadrmum dinitnde
Tncalaum dinitnde
Tncesium nitride
Tnchloroacetaldehyde
Tnchloroborane
Tnchloroethane
Tnchloroethene
Tnchloroisocyanunc acid
Tnchloromethane
Tnchloromethyl sulfenyl chlonde
Tnchloronitromethane
TnchlorophenoxyaceUc acid
Tnchloropropane
Tnchlorosilane
Tndecene
Tnethanolamine
Tnethyl aluminum
Tnethyl antimony
Tnethyl arsine
Tnethyl tesmuthine
Tnethylamine
Tnethylene phosphoramide
Tnethylene tetraamine
Tnethyl stibine
Tnfluoroethane
Tnfluoromethylbenzene
Tmsobutyl aluminum
Tnlead dinitnde
Tnmercury dinitnde
Tnmethyl aluminum
Tnmethylamine
Tnmethyl antimony
Tnmethyl arsine
1 ,2,4-Tnmethy Ibenzene
1 ,3 ,5-Tnmethylbenzene
Tnmethyl bismuthine
Tnmethyl pentane
Tnmethylstibine
Tn-n butylborane
Tnnitroaniline
Tnmtroanisole
Tnratrobenzene
Tnmtrobenzoid acid
RGN
24,25
25
24,25
5,17
107
17
17
104
17
17,20
17, 27, 102
3,17
17
107
28
4.7
105,107
24, 105, 107
24,107
24
7
6,32
7
24, 105, 107
17
17
105, 107
24,25,102
24, 25, 102
105, 107
7
24,105
24,107
16
16
24
29
24, 105, 107
105, 107
7,27,102
14,27
27,102
3,27,102
D-26
-------
FIGURE D.2 (Cont.)
RGN'S FOR SELECTED MATERIALS
Names
T
RGN
Names
RGN
Tnmtroglycenn
Truutronaphthalene
Tniutrophenol
Tnnitrophenyl methyl ether
Tnnitroresorcinol
Tnmtrotoluene
Tnoctyl aluminum
Tnphenyl ethylene
Tnphenyl methane
Tnpropylamine
Tnpropyl subine
Tnsilyl arsine
Tns-(l-azindinyl) phosphine oxide
Tnthion
Tnthonum tetiarutnde
Tnvmyl subine
Tsumacide*
Tungstic acid
Turpentine
UDMH
Ultraade*
Undecene
Umsolve
Uranium nitrate
Uranium sulfide
Uranyl nitrate
Urea formaldehyde
Urea nitrate
VC
Valeraldehyde
Valeramide
Valenc acid
Vanadic acid anhydnde
Vanadium oxytnchlonde
Vandadium pentoxide
Vanadium sulfate
Vanadium tetroxide
Vanadium tnchlonde
Vanadium tnoxide
Vanadyl sulfate
Vapona*
Vinyl acetate
Vinyl azide
27,102
27,102
27.31.102
14.27
27.31,102
27.102
105,107
16
16
7
24.107
24.107
6.32
32
24,25
24,107
9
24
101
8
32
28
101
24,104
24, 33. 105
24,104
5
27. 102. 104
17,103
5
6
3
24
24
24
24
24
24.107
24
24
32
13.103
102
Vinylbenzene
Vinyl chloride
Vinyl cyanide
Vinyl etheyl ether
Vinyl isopropyl ether
Vinykdene chloride
Vinyl toluene
Vinyl tnchlorosilane
VX
Water
Waxes
Wepsyn* 155
Wood
Zectran*
Zinc
Zinc acetyhde
Zinc ammonium nitrate
Zinc arsenate
Zine arserute
Zinc chlonde
Zinc dioxide
Zinc ethyl
Zinc cyanide
Zinc fluoborate
Zinc nitrate
Zinc permanganate
Zinc peroxide
Zinc phosphide
Zinc salts of dimethyl dithiocaibamic acid
Zinc sulfate
Zinc sulfide
Zineb*
Zmophos*
Ziram*
Zirconium
Zirconium chlonde
Zirconium picramate
Zirconium tetrachlonde
16,28,103
17,103
26,103
14
17
17,103
28,103
107
20.32
106
101
6,32
101
9
22, 23. 24
24. 105. 107
24,104
24
24
24
24, 102, 104, 107
24. 105. 107
11 24
24,15
24.104
24,104
24. 102. 104. 107
24.107
12.24
24
24. 33, 105
12,24
20
12.24
22,23.24
24
24,104
24
D-27
-------
REFERENCES
1. Armistead, G. Safety in Petroleum Refining and Related Industries. John D. Simmons
and Co., Inc., New York, 1950.
2. Assessment of Industrial Hazardous Waste Practice Series U.S. Environmental
Protection Agency, Office of Solid Waste Management Programs, Washington, D C
Electroplating and Metal Finishing Industries - Job Shops. Battelle Columbus Labora-
tories, 1976.
Inorganic Chemical Industry. Versar, Lie., 1974.
Leather Tanning and Finishing Industry. SCS Engineers, Inc., 1976
Metals Mining Industry. MRI, Inc, 1976
Organic Chemicals, Pesticides, and Explosives Industries TRW Systems, Inc., 1976
Paint and Allied Products Industry, Contract Solvent Reclaiming Operations, and
Factory Applied Coatings. Wapora, Inc., 1976.
Petroleum Refining Industry. 1976.
Pharmaceutical Industry. A D. Little, Inc., 1976.
Rubber and Plastic Industry. Foster D. Snell, L.C., 1976.
Storage and Primary Battenes Industries. Versar, Inc., 1975.
3. Battelle Memorial Institute. Program for the Mangement of Hazardous Waste Vol-
umes 1 and 2 US. Enviornmental Protection Agency, Office of Solid Waste
Management Programs, Washington, D.C., 1974.
4. Bonner, W.A. and A J Castro. Essentials of Modem Organic Chemistry Remhold
Publishing Corporation, New York, 1965.
5. Booz-AUen Applied Research, Inc A Study of Hazardous Waste Materials, Hazardous
Effects, and Disposal Methods U.S. Environmental Protection Agency, Cincinnati
Laboratories, Cincinnati, Ohio, 1972
6. Brescia, F., J. Arents, H Meislich, and A. Turk Fundamentals of Chemistry.
Academic Press, New York and London, 1970
D-28
-------
7. Brethenck, L. Handbook of Reactive Chemicals Hazards. CRC Press, Inc., Cleve-
land, Ohio, 1975.
8. Vector and Waste Management Section Files. California Department of Health;
Berkeley, Los Angeles, and Sacramento, California
9. California Industrial Waste Surveys. California Department of Health, Sacramento,
California, 1977
10. California Liquid Waste Haulers Record California Department of Health, Sacramen-
to, California, 1977
11. California Manufacturers Association California Manufacturers Register. Times Mir-
ror Press, Los Angeles, California, 1975
12. Case Histories of Accidents in the Chemical Industry Manufacturing Chemists
Association, Washington, D C.; 1962, 1070, and 1976. Case History Nos.: 262, 1566,
and 2172.
13 Chemical Data Guide for Bulk Shipment by Water U.S Department of Transporta-
tion, Coast Guard, Washington, D.C., 1976.
14. Chemical Hazards Response Information System, Hazardous Chemical Data Volumes
1 and 2. CG-446-2, U S. Department of Transortation, Coast Guard, Washington,
D C, January 1974.
15 The Chemistry of Acrylonitnle Second Edition. American Cyanamid Co.,
Petro-Chemical Dept, New York, 1959
16 The Chemistry of the Ammo Group. S. Patai, editor. Interscience Publishers, New
York, London, Sydney, Toronto, 1968
17. The Chemistry of the Carbonyl Group. J. Zabicky, editor. Interscience Publishers,
New York, London, Sydney, Toronto, 1970
18 The Chemistry of the Cyano Group Z Rapoport, editor Interscience Publishers, New
York, London, Sydney, Toronto, 1970
19. The Chemistry of the Nitro and Nitroso Groups, Part I H Feuer, editor Interscience
Publishers, New York, London, Sydney, Toronto, 1969
20 The Chemistry of the Thiol Group, Part I S Patai, editor, Interscience Publishers,
New York, London, Sydney, Toronto, 1974.
D-29
-------
21. Concise Chemical and Technical Dictionary. Thud Edition. H. Bennett, editor.
Chemical Publishing Co, Inc., New York, NY, 1974
22. Cotton, FA. and G. Wilkinson. Advanced Inorganic Chemistry. Third Edition.
Interscience Publishers, John Wiley and Sons, Inc., New York, London, Sydney,
Toronto, 1972.
23. CRC Handbook of Chemistry and Physics. 54th Edition R.C. Weast, editor. CRC
Press, Inc., Cleveland, Ohio, 1973.
24. CRC Handbook of Laboratory Safety N.V. Steere, editor. The Chemical Rubber
Company, Cleveland, Ohio, 1967.
25. Encyclopedia of Chemical Technolgoy, Kirk-Othmer. Second Edition. M. Grayson
and D. Eckroth, editors. John Wiley and Sons, Inc., New York, London, Sydney,
Toronto, 1963-1972.
26. Environmentally Hazardous Waste Disposal Files. Oregon Department of Environ-
mental Quality, Portland, Oregon, 1976 and 1977.
27. Epoxy Resins. Advances in Chemistry Series No. 92. R F Gould, editor. American
Chemical Society, Washington, D.C., 1970.
28. Eto, M. Organophosphorous Pesticides. Organic and Biological Chemistry. CRC
Press, Lie, Cleveland, Ohio, 1974.
29. Faith, W.L., D.B. Keyes, and R.L. Clark. Industrial Chemicals Third Edition. John
Wiley and Sons, Inc., New York, London, Sydney, 1966
30. Farm Chemicals Handbook, 1978. Meister Publishing Company, Willoughby, Ohio,
1978.
31. Fieser, L J. and M. Fieser Advanced Organic Chemistry Remhold Publishing Corpo-
ration, New York, 1961.
32. Fire Protection Guide on Hazardous Materials. Sixth Edition National Fire Protection
Association, Boston, Massachusetts, 1975.
33. The Form of Hazardous Waste Materials Rollins Environmental Services, Inc.
Wilmington, Delaware, 1972
34. Gardner, W. and EE. Cooke Chemical Synonyms and Trade Names Seventh
Edition. CRC Press, Inc., Cleveland, Ohio, 1971.
D-30
-------
35. Geissman, T.A Principles of Organic Chemistry. WH. Freeman and Company, San
Francisco, 1977.
36 Gorham International, Inc. Study of Solid Waste Management Practices in the Pulp
and Paper Industry US. Environmental Protection Agency, Washington, D.C.,
February 1974.
37. Guide for Safety in the Chemical Laboratory, Manufacturing Chemists Association
Second Edition. Van Nostrand Rheinhold Co., New York, NY 1972.
38 Handling Guide for Potentially Hazardous Materials. A D Baskm, editor. Material
Management and Safety, Inc., Niles, IL, 1975.
39. Handling and uses of the Alkali Metals. Advances in Chemistry Series No. 19,
American Chemical Society, Washington, D.C., 1957.
40. Hawkins, E.G.E. Organic Peroxides. D. Van Nostrand Company, Inc., Toronto, New
York, London, 1961.
41. Hawley, G.G. The Condensed Chemical Dictionary. Eight Edition Van Nostrand
Reinhold Company, New York, Cincinnati, Toronto, London, Melbourne, 1971.
42. Hazardous Waste Disposal Damage Reports U S. Environmental Protection Agency,
Office of Solid Waste Management Programs, Hazardous Waste Management Divi-
sion, Washington, D.C., June 1976.
43. Hazardous Waste Management Law, Regulations, and Guidelines for the Handling of
Hazardous waste. California Department of Health, Sacramento, California, February
1975.
44. Hendnckson, J.E., DJ. Cram, and G S. Hammond. Organic Chemistry Third Edition.
McGraw-Hill Book Co., New York, 1970.
45. House, H.O. Modern Synthetic Reaction. W A Benjamin, Inc., Memo Park, Califor-
nia, 1972.
46. Industrial Hygiene and Toxicology. Volumes I-m. F.A Patty, editor, Interscience
Publishers, Inc., New York, 1958.
47. Industrial Pollution Control Handbook. HF. Lund, editor. McGraw-Hill Book Co.,
New York, St. Louis, San Francisco, 1971.
D-31
-------
48. Kimball, V.S. Waste Oil Recovery and Disposal. Noyes Data Cororation, New Jersey,
London, 1975.
49. Kuhn, R.J. and H.W. Dorough. Carbamate Insecticides: Chemistry, Biochemistry, and
Toxicology. CRC Press, Inc, Cleveland, Ohio, 1976.
50. Lawless, E.W, TJL. Ferguson, and A.F. Memers Guidehnes for the Disposal of Small
Quantities of Unused Pesticides U.S. Environmental Protection Agency, Office of
Research and Development, National Environmental Research Center, Cincinnati,
Ohio, June 1975.
51. Margenson, D. and G L East Introduction to Polymer Chemistry. Pergamon Press,
Ltd., Oxford, London, New York, 1967.
52. MC/B Chemical Reference Manual. Volume U. MC/B Manufacturing Chemists,
Norwood, Ohio, June 1976.
53. Millan, I Ketones. Chemical Publishing Co., New York, 1968.
54. The Merck Index. Ninth Edition Merck and Company, Inc., Rahway, New Jersey,
1976.
55. Morrison, RT. and R.N. Boyd. Organic Chemistry. Second Edition. Allyn and
Bacon, Inc., Boston, 1969
56. Nemerow, N.L. Liquid Waste of Industry: Theories, Practice, and Treatment Addi-
son-Wesley Publishing Co., Reading, MA, Menlo Park, CA, London, Don Mills,
Ontario, 1972.
57. An Outline of Organic Chemistry, College Outline Series. F. Degermg, editor Barnes
and Noble, Inc., New York, 1946.
58. Osmond, ROD. Personal Communication. Department of the Environment, London,
England, April 1977.
59. Pesticide Index. Fifth edition. WJ Wiswesser, editor. The Entomological Society of
Americal, College Park, Maryland, 1976.
60. RUder, L.R., JH. Cobbs, JW. Fields, Jr., WD. Fmdley, SL. Vokurka, and B.W.
Rolfe. Review and Assessment of Deep Well Injection of Hazardous Waste. U.S.
Environmental Protection Agency, National Environmental Research Center, Solid and
Hazardous Waste Research Laboratory, Cincinnati, Ohio, 1975.
D-32
-------
61. Registry of Toxic Substances. 1976 Edition HE. Chnstensen and EJ Fairchild,
editors U.S Department of Health, Education, and Welfare, Rockville, Maryland,
June 1976.
62 Report to Congress* Disposal of Hazardous Wastes. U.S Environmental Protection
Agency, Office of Solid Waste Management Programs, Washington D C., 1974.
63. Resource Conservation and Recovery Act of 1976 PL 94-580,94th Congress, October
21,1976
64 Reigel's Handbook of Industrial Chemistry Seventh Edition J A Kent, editor. Van
Nostrand Reinhold Company, New York, Cincinnati, Toronto, London, Melbourne,
1974.
65. Rmehart, K L. Oxidation and Reduction of Organic Compounds Prentice-Hall, Inc,
Englewood Chffs, New Jersey 1973
66 Royals, EE Advanced Organic Chemistry. Prentice-Hall, Inc, Englewood Cliffs,
New Jersey 1959.
67 Rutledge, TF Acetylenes and Allenes. Remhold Book Corporation, New York,
Amsterdam, London, 1969.
68. Sandier, S.R. and W. Karo. Polymer Synthesis. Volumes I and II. Academic Press,
Inc., New York, 1974.
69. Sax, I.N. Dangerous Properties of Industrial Materials Third Edition. Van Nostrand
Reinhold Company, New York, 1968.
70. Schieler, L. and D. Pauze. Hazardous Materials. Delmar Publishers, Albany, New
York, 1976.
71 Sidgwick, N V. The Organic Chemistry of Nitrogen. Clarendon Press, Oxford, 1966.
72 SRI, International, Handbook of Hazardous Waste. Federal Ministry of the Interior,
Federal Republic of Germany, 1974.
73. Standard Industrial Classification Manual. Executive Office of the President, Office of
Management and Budget, Statistical Policy Division, Washington, D.C, 1972.
D-33
-------
74. Stone, R.B., PJL. Aamedt, M.R. Engler, and P. Maiden. Evaluation of Hazardous
Waste Emplacement in Mine Openings U.S. Environmental Protection Agency,
Muncipal Environmental Research Laboratory, Office of Research and Development,
Cincinnati, Ohio, December 1975.
75. Szmant, H.D. Organic Chemistry. Prentice-Hall, Inc., Englewood Cliffs, New Jersey,
1957.
76. Toxic and Hazardous Industrial Chemicals Safety Manual for Handling and Disposal,
with Toxicity and Hazard Data. The International Technical Information Institute,
Toranomon-Tachikawa Bldg. 6-5, 1 Chome, Nishi-Shimbashi, Minato-Ku, Tokyo,
Japan, 1975.
77. TRW Systems, Inc. Recommended Methods of Reduction, Neutralization Recovery,
or Disposal of Hazardous Waste Volumes I-XVL U.S. Environmental Protection
Agency, Washington, D.C., 1973.
78. Wiberg, E., and E. Amberger. Hydrides of the Elements of Main Groups I-IV.
Elsevier Publishing Co., Amsterdam, London, New York, 1971.
79. Zollinger, H. Azo and Diazo Chemistry: Aliphatic and Aromatic Compounds. Inter-
science Publishers, Inc., New York, London, 1961.
D-34
-------
APPENDIX E
GUIDE TO INSTALLATION OF THE ARCHIE COMPUTER PROGRAM
Introduction
The purpose of this appendix is to provide assistance with installation of the Automated
Resource for Chemical Hazard Incident Evaluation (ARCHIE) computer program Due to
the very unusual nature of the installation procedure, it is highly recommended that all users
read and follow subsequent instructions as closely as possible. It is also suggested that
introductory portions of Chapter 12 be read prior to program installation up to and including
the section on system initialization.
System Requirements
ARCHIE may be installed on IBM™ personal computers (PCs) and those other such
computers that are fully compatible with IBM products and which operate under PC-DOS or
MS-DOS disk operating systems of Version 2.0 or higher (with certain exceptions noted
below) Hardware requirements include:
• A monochrome or color 80-column monitor linked to a monochrome
display adaptor (MDA), color graphics adaptor (CGA), enhanced graphics
adaptor (EGA), or video graphics array (VGA) video display card.
• Two disk drives, at least one of which can read 5 25 inch double-sided,
double density, floppy diskettes of 360k capacity.
• 512k of free random access memory (RAM)
• A compatible printer if written summaries are desired
Special Self-Extracting Program Feature
ARCHIE and its associated program files do not normally fit within the space available
on a single 360k diskette, yet this is currently the most common type of disk in use and
therefore the best choice for widespread distribution of the program In order to fit the
program on one disk, and thereby save substantial duplication and distribution costs for a
second, the developers of ARCHIE licensed use of an unusual and possibly unique program
utility from a pnvate vendor. This utility is unusual in that it permitted the author of
ARCHIE to compress many program files to a fraction of their normal size and then pack
them together into just the two files that appear on the distribution disk. The utility is rather
E-l
-------
unique in that a single command will cause each resulting Hie to unpack and decompress
itself while automatically directing decompressed files to the drive and/or path specified by
the end user. The original packed files are not altered or modified in any way by this
process. Thus, the user may start over at any time if a mistake is made during program
installation.
The two packed files provided on the diskette received with this document are
respectively named VOLUME1.EXE and VOLUME2.EXE. How one goes about unpacking
and decompressing them for final use depends on the types of disk drives installed in the
user's computer system and is discussed in the next section.
Note that the following instructions assume that the user has at least minimal
knowledge of file management techniques on personal computers. Where instructions are
not understood, please refer to the manuals that accompanied your DOS diskettes Key
words to look for in these manuals are shown in italics.
Program Installation Instructions
The following instructions should be adequate to install ARCHIE on the majority of
IBM and compatible personal computer systems. Users who encounter difficulty should
consult a more experienced computer user for assistance.
For those of you with fairly new computers having only 3.5 inch floppy drives, it will
be necessary to copy the ARCHIE program files to a 3.5 inch disk pnor to installation Since
most of you will have encountered this type of disk conversion problem before, it is expected
that relatively few users will expenence insurmountable problems in undertaking this task
For Systems with a Hard Drive or Card:
Assuming that the computer will boot automatically (i.e., show a system prompt) when
turned on, the best approach is to install the entire contents of the program diskette on the
hard drive or card in its own separate directory Once all files have been unpacked,
decompressed, and copied to this directory, and this directory is the current directory, it is
only necessary to type "ARCHIE" followed by a press of the ENTER key to start the
program.
A directory is created on a hard disk, when the user in the root directory, by typing
"MDNDIRNAME" (leaving out the quote marks shown here and in other commands that
follow), where DIRNAME can be any one to eight-letter name of the users choosing The
user can enter or proceed to the new directory (and thereby make it the current directory) at
any time by typing "CDXDIRNAME" when the hard drive is the default drive (i e., when
"C:>" appears on screen as the system prompt)
E-2
-------
With the ARCHIE distribution disk in drive A, the user should next type "A:" and press
ENTER so that drive A becomes the default drive with "A:>" appearing as the system
prompt on the screen. It then becomes time to actually unpack and decompress the ARCHIE
program files, this being accomplished by typing:
• "VOLUME1 C:\DIRNAME" followed by a press of the ENTER key
• " VOLUME2 C:\DIRNAME" followed by a press of the ENTER key.
As the program files unpack and decompress, numerous messages will be shown on
screen. It is prudent to wait until VOLUME1 has decompressed and the system prompt has
reappeared before giving the second command listed above.
While VOLUME2 is decompressing, the user will be asked whether or not existing
files should be overwritten. The answer that should be given in response does not make a
difference, but a no answer will save a few seconds of file transfer time. When the system
prompt reappears.
Type "C:" followed by a press of the ENTER key.
Type "CDOIRNAME" followed by a press of the ENTER key.
Type "ARCHIE" followed by a press of the ENTER key to start the
program.
• See Chapter 12 to learn more about what happens next.
A viable alternative for those who do not wish to copy the program onto then: hard
drive or card is to follow the directions given below for systems with two floppy drives with
two possible modifications. The first is that the diskette(s) to which the decompressed files
are to be directed may not need to be formatted with the 7S" suffix. The second is that the
user has a choice of either drive B (if present) or drive C as storage locations for the Accident
Scenario Files (ASF) described in Chapter 12
For Systems With Two 360k Floppy Drives and No Hard Drive
If your computer has two 5.25 inch floppy drives of 360K capacity, follow these
instructions.
• Format two blank diskettes as you normally would, but add the suffix "/S"
to the format command. For example, with your DOS diskette in drive A,
and an unformatted diskette in drive B, type "FORMAT B:/S".
E-3
-------
• If the system prompt is not "A:>"at this time, type "A." followed by a press
of the ENTER key.
• Place one of the newly formatted diskettes in drive B and place the
ARCHIE distribution diskette in drive A.
Type "VOLUME1 B:" followed by a press of the ENTER key and wait for
the system prompt to reappear after various messages.
Remove the disk from drive B and label it as "ARCHIE Volume I". This
disk will contain the first of two sets of ARCHIE program files and can be
used to boot the system.
• Place the second newly formatted diskette in drive B.
• Type "VOLUME2 B:" followed by a press of the ENTER key and wait for
the system prompt to reappear after various messages.
• Remove the disk from drive B and label it as "ARCHIE Volume II". This
disk will contain the second of two sets of program files.
• Remove the DOS diskette from drive A and place the ARCHIE Volume I
disk in this drive.
Type "ARCHIE" followed by a press of the ENTER key to start the
program. See Chapter 12 for what happens next.
For Systems With Two Floppy Drives of Different Capacity and No Hard Drive
If your computer has two floppy drives, one of which is of 360K capacity, and the
other of which is of higher capacity (and therefore uses high density diskettes), it is assumed
that the higher capacity drive is drive A, since this is the most common system configuration.
Follow these instructions to place both sets of ARCHIE program files on one high density
diskette together with the operating system files.
• Format one blank high density diskette as you normally would, but add the
suffix 7S" to the format command For example, with your 360k DOS
diskette in drive B, and an unformatted high density diskette in drive A,
type "B:FORMAT A:/S".
• If the system prompt is not "B:>"at this time, type "B:" followed by a press
of the ENTER key.
E-4
-------
Place the newly formatted high density diskette m drive A and place the
ARCHIE distribution diskette in drive B.
Type "VOLUME1 A." followed by a press of the ENTER key and wait for
the system prompt to reappear after various messages
Type "VOLUME2 A-" followed by a press of the ENTER key While
VOLUME2 is unpacking and decompressing, expect to be asked whether or
not existing files should be overwritten The answer that should be given in
response does not make a difference, but a no response will save a few
seconds of file transfer time.
When the system prompt reappears, remove the high density disk from
drive A, label it as "ARCHIE", and replace it in drive A Remove your
DOS diskette from drive B. The single high density diskette will contain
all program files of ARCHIE plus sufficient room for quite a few Accident
Scenario Files (ASF). The disk can be used to boot the system
Type "A-" followed by a press of the ENTER key to make drive A the
default drive. Type "ARCHIE" followed by a press of the ENTER key to
start the program, and see Chapter 12 for what happens next.
Additional Instructions For any System With a Color Monitor
In order for ARCHIE to display certain screens in full color, it is necessary for the
ANSIS YS driver to be installed in the CONFIG SYS file of the computer. For those readers
that do not understand the above statement, follow these instructions:
Locate the file named ANSI.SYS on your DOS diskettes.
Copy this file to the floppy diskette or root directory of the hard drive from
which the system is booted up when power is turned on.
• Look at the above diskette or location using the DIR command to determine
whether a file with the name CONFIG SYS exists
If CONFIG.SYS does exist, add the line "DEVICE = ANSI.SYS" without
quotes to the list of commands in the file Modify the file using the copy
con command described in your DOS manuals, the edhn editor provided on
your DOS diskettes, or any other ASCII file editor
E-5
-------
• If CONFIG.SYS does not exist, see your DOS manuals for the purpose of
this file and its contents. Create the file using the copy con command
described in your manuals, the edhn editor provided on your DOS
diskettes, or any other ASCII file editor.
Potential Problems that May be Experienced
In order to properly unpack and decompress ARCHIE, the self-extracting program
utility requires access to specific portions of the computer's memory. This access may at
times be impeded or interrupted by TSR (terminate and stay resident) programs that may
have been installed earlier in memory. Thus, if problems are expenenced during file
decompression or initial use of ARCHIE, it may be necessary to reboot the computer (or
otherwise ensure that TSR programs have not be installed) and repeat the installation
procedure. (Note: A TSR program is usually one that can be activated at any time by the
user, even while he or she is in the middle of another program, by pressing a specific
combination of keys. Alternatively, it may be a performance enhancement utility of some
type loaded into memory for one reason or another) Users in office environments who
experience problems and are not sure whether or not their system automatically installs one
or more TSRs should seek assistance from a more expenenced computer user — who is
advised to check the contents of the AUTOEXEC.BAT file on the boot disk or drive for
indications that a TSR program is being installed
Some versions of DOS 2.X have been identified as having difficulties executing an
important and frequently used instruction in ARCHIE referred to as a "SHELL" command,
even though their documentation has always claimed that use of this instruction is fully
acceptable. If problems are expenenced during program use, it may be necessary to obtain
and install DOS version 3 0 or above in order for ARCHIE to function properly. (Note:
Typical manifestations of this problem produce "Illegal Function Call" and "Wrong Version
of Command.Com" error messages during program use. Although special steps have been
taken to avoid the problem and it should arise infrequently, some users may nevertheless
experience difficulties).
Users who have a color monitor, and who neglect to install the ANSI SYS driver as
instructed above, may not be able to proceed past the second screen of the program
initialization procedure discussed in Chapter 12 The presence of this driver is mandatory if
the user specifies that a color monitor is present
E-6
-------
APPENDIX F
BASIS OF PROBABILITY ANALYSIS PROCEDURES
Introduction
This appendix provides and discusses the basis for the probability analysis procedures
presented in Chapter 11 of the guide.
Transportation of Hazardous Materials by Highway
Available Accident Data for Bulk Transportation by Highway
Although numerous data bases and compilations of spill records are available for
highway transportation, the purpose of many of them is such that they are not conducive to
obtaining accident frequency estimates, primarily because they keep track of the number of
events but not the total exposure Such data bases can still be useful, however, in determining
breakdowns of accidents by size or cause Available data bases and their completeness, as
noted above, are described by both the July 1986 OTA report and a series of reports prepared
by Midwest Research Institute for the Federal Highway Administration (1987a and b) This
section presents some of the failure rates and other statistics given in the literature Values
actually recommended for use in emergency planning follow in a later section
The rate of accidents can be a function of road type (urban, rural, etc), number of lanes,
traffic density, average speeds, type of vehicle, number of intersections, road conditions,
weather conditions, geometry of the road, grade, etc However, differences attributed to these
various causes tend to give results that are within roughly one order of magnitude, with the
range usually being 1 to 10 x lOVmile or between one and ten accidents per million miles
driven. (Urbanek and Barber 1980, API 1983, Smith and Wilmot: 1982, National Safety
Council: 1988).
Frequently cited average accident rates in prior studies are
• 5 0 x lO-Vmile for trucks in the petroleum industry (API, 1983)
2 5 x 10 6/mile for trucks (Dennis et al. 1978, Rhoads et al 1978 and others)
• 1.4 x 10-Vmile for trucks with trailers in California (Smith and Wilmot, 1982)
• 12x10 s/mile for all trucks, and
3 1 x 10-Vmile for intercity trucks (National Safety Council, 1988)
F-l
-------
• 1.5 x KWmile for bulk hazardous materials trucks (Ichmowski, 1984)
8.3 x 10-Ymile for trucks (Kloeber et al, 1979)
0.9 - 2.1 x lOVmile for single trailer trucks (DHS, 1985)
Yet other rates have been reported for specific locations or road types Much of the variation
in these average rates can be explained by level of compliance with reporting requirements and
different reporting thresholds in terms of damages sustained for the various data bases, as well
as the road and weather conditions in the subject area (Note: See Appendix A if you are not
famihar with scientific notation of numbers such as 5 x 1O6 )
With respect to the fraction of all reported accidents that result in a spill or discharge, the
range of estimates in the literature include 0 20 (IGF, 1984), 0.115 (OTA, July 1986), 0 30
(Elder et al, 1978), 0.46 (ADL, 1979), and other values which range from less than 0 01 to
about 0.50. One source states that 0.3-1.2% (0.003-0 012) of most types of truck accidents
result in a fire (Dennis et al, 1978) Some data sources combine the accident rate with
prespecified levels of accident seventy, for example (Clarke et al, 1976)
Minor 2 4 x
Moderate 4 5 x
Severe 7.2 x lO-Ymile
Extra severe 3 5 x lO^/rmle
Extreme 1.2 x 10 9/mile
A review of hazardous material accidents on highways over the five-year period 1981
through 1985 was earned out by MRI (1987). This study concluded that, based on truck
accidents reported to the Bureau of Motor Gamer Safety (BMCS) of the Federal Highway
Administration, 152 percent of accidents involving hazardous material-carrying vehicles
resulted in a release. Accidents involving tank trucks resulted in releases 16 6 percent of the
time based on 1984-1985 BMCS-reported accident data. It is not clear whether accidents
involving empty trucks which normally carry hazardous matenals were included in the data
base; the implication in this study, however, is that only loaded trucks are included
The distnbution of spill amounts in accidents involving spills is somewhat less variable
One study of data in the Pollution Incident Reporting System (PIRS) of the U.S. Coast Guard
found an average spill volume of 1880 gallons for highway transport spills of hazardous
matenals. An analysis of LPG truck accidents (Croce et al, 1982) gave breakdowns as follows
for small (<7000 gal) and large (>7000 gal) trucks:
F-2
-------
Spill Size
<100gal
100 - 1,000 gal
1,000 -5,000 gal
5,000 -10,000 gal
Small Trucks
49%
18%
27%
6%
Large Trucks
44%
19%
12%
25%
Another study found an average truck release volume of 2900 gallons (excluding spills of less
than 100 gal), with 23% being less than 100 gal, 22% from 100 - 500 gal, 8% from 500 -1000
gal, 16% from 1000 - 5000 gal, and 31% from 5000 to roughly 10,000 gal (ADL, 1978).
Rhoads et al (1978) found that about 7% of all spills from gasoline trucks exceed 3000 gallons.
The Hazardous Material Information System of the U S Department of Transportation
has maintained a data base on the size of reported releases of LPG For the eleven-year penod
of 1976-1986,76 releases of LPG were reported in highway accidents. These were distributed
as follows:
Small releases (1 -1000 gallons)
Large releases (> 1000 gallons)
= 53 percent
= 47 percent
Use of Local Data
State and local highway departments may maintain accident counts or rates, but it is less
likely that they will have information on commodity flows. For most types of accidents, it is
probably desirable to stay at the level of state or regionwide data in order to avoid excluding
very low frequency events. Nationwide data on accident rates is probably a reasonable
approximation, and the focus should be on improving the commodity flow data base rather
than on the accident data base if only one can be addressed However, whenever local or state
data is available, it can be substituted for the values used m the worksheet given in Chapter 11.
The City of Portland, Oregon, is an example of a community that has conducted a
detailed analysis of the materials spilled in past accidents and the types of accidents causing
these releases. It found that 24 5% of truck accidents resulted in spills. From 1976-1980,
gasoline accounted for 47.3% of spills, diesel 23%, asphalt 8.2%, fuel oil 4.1% and other
materials 17.4%. Commodity flow data were also obtained, but it was acknowledged that this
information was somewhat incomplete (Robison, 1981). Table F.I summarizes statistics on
truck accidents gathered for the state of Pennsylvania in recent years.
F-3
-------
TABLE F.I
PENNSYLVANIA HIGHWAY ACCIDENTS
._J_ - -- -t 1-- - -' •'• • ~.'^~~ .. .1 TIMIT.r. • 1. • -~ TlT
Accidents reported
Fatalities
Injured
Caused by drivers with cargoes of hazardous materials
Caused by other drivers
Other causes
Number of hours highways were restricted
Number of hours highways were closed
Accidents resulting in loss of product
Accidents resulting in fire
Accidents involving:
Explosives and blasting agents
Flammable liquids and solids
Oxidizing materials, poisons, and corrosive materials
Radioactive materials
Compressed gases - cryogenics
Combustible liquids
Percentage of accidents involving cargoes of:
Flammable liquids and solids, %
Compressed gases - cryogenics, %
Explosives and blasting agents, %
Radioactive materials, %
Oxidizing materials, poisons, and corrosive materials, %
- -- - -
1980
174
24
73
89
82
5
137
140
33
4
2
87
20
3
16
46
50
9
1
2
12
26
1981
165
9
200
81
62
10
77
164
31
1
6
67
22
1
20
49
41
12
4
1
13
29
—
1982
274
11
257
86
88
88
96
146
47
4
6
100
43
3
17
106
36
6
2
1
16
39
- - .-
1982
186
16
164
55
75
56
112
246
27
4
1
73
20
0
32
73
37
16
1
0
10
37
1984
207
14
184
48
76
83
110
110
30
0
2
91
25
0
29
76
41
13
1
0
11
34
Source. Skolmck 1986
-------
Suggested Approach for Assessment of Accident Potential
Since we are concerned with accidents with the potential to cause major problems for a
community or other jurisdiction and not those which are handled on a routine basis, it is best to
focus on vehicular accidents rather than relatively minor leaks from valves, fittings, or open
relief valves Based on the information presented above, an average accident rate of 2 x 1O6
accidents/mile is considered representative of the general experience of trucks carrying bulk
quantities of hazardous materials. If adequate local/state data are available for determination
of individual accident rates for divided and undivided roadways, their use is recommended
because the resulting rates will more accurately reflect accident probabilities under local
conditions.
With respect to the fraction of truck accidents that result in a spill or discharge, the
available data suggest a consensus opinion on the order of 0 50 (50%) if all spills including
very minor valve and fitting leaks are considered Omitting these, a spill appears to result from
an accident m about 0 15 - 0.20 (15 - 20%) of accidents A value of 0.20 (20%) is therefore
suggested for the sake of conservatism.
Based upon the available spill amount distributions, and considering the likely causes of
accidents, the following distribution is suggested for general use*
10% cargo loss (thru 1" hole) or 1000 gal --60% of the time
30% cargo loss (thru 2" hole) or 3000 gal --20% of the time
100% cargo loss (instantly) or 10,000 gal — 20% of the time
These values cover the range of significant releases. If desired, a two-point distribution
assuming 3000 gallon spills 80% of the time and 10,000 gallon spills 20% of the time may be
used to simplify accident consequence estimation procedures.
Transportation of Hazardous Materials by Rail
Available Accident Data for Bulk Transportation by Rail
The overall accident rate for railroads has been reported as being 4 6 x 10* accidents per
train-mile traveled in 1987. This rate was comprised of 4.9 x 10-7 collisions per train-mile, 3.2
x 10* derailments per train-mile, and 8.6 x 10-7 other types of accidents per train-mile. The
general trend has been a reduction in the overall accident rate, the collision rate, and the
derailment rate, with only the rate for "other" accidents holding at about one per million
train-miles (FRA, 1988), as might be expected due to the many new regulations adopted in
recent years to improve railroad safety. For example, the overall accident rates reported for
-------
Year
1984
1983
1982
1981
1980
1979
Accidents per
million train-miles
6.6
7.0
8.0
8.6
11.8
128
(Note: Some adjustments were made in the rates to account for changes in reporting
thresholds.) This compares to the rate of 4.6 per million train-miles in 1987.
There were 54 railroad accidents in 1984 in which hazardous materials were released.
The number of cars that released hazardous materials in these 54 accidents was 100, with 89 of
these being involved in derailments Collisions accounted for about 11% of the 54 accidents,
with derailments accounting for another 75% (ERA, 1985). Over 90% of releases have been
attributed to derailments in the past (von Herberg, 1979), so design improvements may be
having a beneficial effect.
The overall rate of 4.6 x 10* accidents per train-mile can be split into a rate of about 2 9
x 10* per train-mile for mainline track and 1 3 x H>5 per train-mile for rail yards (FRA, 1988).
For a recent five-year period, the average number of cars per freight train has been about 70
(AAR, 1985), and the number of cars involved in each accident has been estimated as between
ten and twenty percent of these on average Trains with hazardous materials cars that
experienced accidents in 1984 contained a total of 2826 cars with these materials. Of these, 581
were damaged and 100 actually released some part or all of their cargo (FRA, 1985) Statistics
for 1982 and 1983 give comparable results (FRA, 1983 and 1984), with the overall fraction of
cars being damaged being in the range of 0.21-0.29 (21-29%) and the fraction of these actually
releasing cargo being in the range of 0 11-0 20 (11-20%) Data for 1987 give a value of 0 18
Harvey et al (1987) report that the trend has remained constant at about 0 17, for the last
several years.
Rates reported in other studies typically reflect accident experience for a year pnor to
that in which the study was completed (or some average of the preceding years), with many
presenting or using an average annual accident rate of 10-5 accidents per train-mile One
particular study found that the chance of an accident resulting in a hazardous cargo release was
2 x 10-7 per car-mile for cars conveying Liquefied Petroleum Gases ~ LPG for short (Kot et al,
1983). Another reported a rate of 4 x 10-7 spills per loaded car-mile using 1976 data (Chemical
F-6
-------
Week, March 15,1978), while a third reported the chance of a fire occurring as 2 8 x 1O8 per
car-mile (Geffen et al, 1980). Note, however, that many of the rates per car-mile that appear in ]
the literature were obtained by dividing the overall accident rate for trains on a mileage basis
by roughly 70 cars per train, without accounting for the fact that multiple cars are generally
involved in each accident. Thus, such rates must be used with care.
Spill amount distributions for rail accidents are provided by only a limited number of
sources. One analyzed 130 accidents to develop the following distribution (Nayak et al, 1983):
Gallons
Less than 1,000
1,000- 5,000
5,000 - 10,000
10,000 - 25,000
25,000 - 50,000
50,000 -100,000
Fraction of Spills
0 37 (37%)
0 12 (12%)
0.12(12%)
0 22 (22%)
009(9%)
008(8%)
Although no single tank car can contain so much, spills of 50,000 gallons or more are reported
above because more than one car may spill its contents in an accident The average spill size
has also been reported as being 11,100 gallons, with 27% of spills being found to be less than
1,000 gallons, 13% being between 1,000 and 5,000 gallons, 25% between 5,000 and 10,000
gallons and the rest being over 10,000 gallons (ADL, 1978).
Use of Local Data
General accident rates for rail transportation can usually be obtained for individual states
and even for individual railroad companies. It is also possible to obtain commodity flow data
on this basis, or even at a more local level by contacting the company operating the track of
concern For example, the City of Portland, Oregon found the following commodity
breakdown based on number of railcars in a fairly recent effort (Robison, 1981)-
Flammable liquids
Flammable compressed gases
Oxidizers
Corrosives
Mixed loads
Non-flammable compressed gases
Class B poisons
311%
159%
146%
117%
115%
94%
31%
F-7
-------
Other 0.8%
Flammable solids 0 7%
Combustible liquids 0 7%
Explosives A,B,C 0.3%
Radioactive material 0 02%
The Portland researchers point out that this list may over-estimate flammable liquids and
underestimate both toxic and flammable liquefied compressed gases.
Harvey et al (1987) give accident rates by material for several materials, ranging from
O.SxlO^/car-mile for chlorine to 9.6xlO*/car-mile for hydrochloric acid. The latter material is
thought to have such a high rate in part due to ruptured frangible discs and failure of the rubber
lining in the car.
Total mileage of track (excluding yards and sidings) is readily available by state (AAR,
1985) and should also be fairly easy to obtain on a subregional or community basis. In such
determinations, the mileage within rail yards will be important not so much in terms of total
track available, but rather in terms of total miles travelled by any one train or car on average
within the yard. Where local data exist as accident rates or spill frequencies, they can be
directly substituted for the values given in the next section
Suggested Approach for Assessment of Accident Potential
Based on the data presented above, it is suggested that an accident rate of 3 x 10* per
train-mile be used for mainline track To convert this to a per car-mile basis, it is assumed that
0.20 (20%) of the cars will be damaged in an accident (based upon data presented in Nayak,
1979). The overall rate therefore becomes 0 2 x 3 x 10-6/ train-miles or about 6 x 107 per
car-mile.
The accident rate for rail yards is obtained by taking 1 3 x 10-5 accidents per train-mile
and a 20% damage estimate to obtain about 3 x 10*/car-mile for the track in yards Sidings
also pose a risk, but these risks are considered herein to be overshadowed by those associated
with mainline and yard track
It is suggested that 0.15 (15%) of accidents be assumed as resulting in a spill for both
mainline and yard accidents, as no data are available to permit distinctions between these
events.
With respect to the distributions of spill amounts in accidents, the available data suggest
use of*
F-8
-------
• 3,000 gallons or 10% of cargo (thru 2" hole) -50% of the time
• 10,000 gallons or 30% of cargo (thru 2" hole) - 20% of the time
• 30,000 gallons or total loss of cargo - 30% of the time
The higher weighting of the last category partially accounts for the potential for more
than one car to release part of its contents in an accident
Marine Transportation of Hazardous Materials
Available Accident Data for Marine Transportation
The accidents of concern for marine transportation include collisions (moving and while
moored or docked), groundings, and rammings Accident rates may be expressed as per port
call, per transit, per mile, per shipment, or per ton-mile.
An analysis of accident rate estimates for collisions, groundings, and all types of moving
vessel casualties in harbors along with moored casualties has demonstrated that accident rates
derived from different data bases for various harbors were quite similar, with each type of
casualty having accident rate estimates ranging within a single order of magnitude (ADL,
1983). Studies of tank barge casualties by the U.S. Coast Guard (1979) and by the Maritime
Transportation Research Board (1981) have talked 229 collisions, 173 rammings, and 71
groundings from 1972-1976. Although marine traffic data are not available for the same time
period, Corps of Engineers data for 1982 permit computation of the following overall accident
rates on a per transit basis:
Collision 3 x 10-ytransit
Ramming 7 x 10-Ytransit
Grounding 3 x 10^/transit
Assuming an average top length of 200 miles, a conservative collision rate in the Gulf
Intercoastal Waterway, which is representative of the Intercoastal Waterway and other
relatively narrow waterways and rivers, can be estimated as being on the order of 5 x 10^
collisions/mile. For more highly traveled inland waterways and major rivers, increased traffic
levels, higher speeds, and larger tows would be expected to increase the rate of collisions.
The Houston-Galveston Vessel Traffic Service (operated by the U.S. Coast Guard) has
compiled statistics showing an annual collision rate within the Service area ranging from 1.5 x
10-4 to 4.7 x 10-» per transit in the period 1977 to 1984, with a mean value of 2 8 x 1O4
collisions/ transit. Studies of marine casualty rates for conventional self-powered ships and
tankers found a grounding casualty rate in the range of 5 x 104 to 4 x 10-s per transit. Moving
collisions ranged from 6 x 10-* to 5 x 10-5 per transit and collisions while moored ranged from
F-9
-------
1-7 x 10s per port call (ADL, 1983) A study by Todd Shipyards (1976) for the
Houston-Galveston port system over the five-year period of 1970-1974 calculated a moored
casualty rate of 1.8 x 104 collisions/port call
The fraction of accidents that result in a discharge were found by the above studies to
range from a low value of about 0.01 (1%) for groundings in deep water and collisions
involving double hulled vessels to a high of about 0 40 (40%) for collisions involving vessels
with single hulls. As reported by Abkowitz and Galarraga (1986), one study found that the
percentage of accidents that results in spills does not vary considerably for different causes of
accidents. This study suggests that about 0.15 (15%) of the accidents of concern result in spills
or discharges.
With respect to the expected volumes of spills, the Coast Guard's Pollution Incident
Reporting System (PIRS) provides data indicating the following average spill amounts during
1983:
Commodity
Chi
Hazardous substances
Other polluting substances
Total
Tank ships
543 gal
108 gal
152 gal
492 gal
Tank barges
3309 gal
4301 gal
84 gal
3275 gal
These spill volumes are lower, however, than the averages found in other studies. Abkowitz
and Galaraga (1986) have reported average spill size by location and given:
Piers
Harbors
Entrances
Coastal
25,000 gal
47,000 gal
115,000 gal
416,000 gal
Other analyses have found that while more than 70% of all spills are less than 100 gal
and 95% or so are less than 10,000 gal (ADL, 1978 and Stewart and Kennedy, 1978), the
average spill volume from a marine vessel is on the order of 44,500 gallons (ADL, 1978). One
reason for discrepancies among studies is that spills of several million gallons can and do occur
on occasion and have the effect of distorting average values for the time periods in which they
occur.
F-10
-------
Use of Local Data
Local data are available from a variety of sources and can be useful for increasing the
accuracy and specificity of accident rate estimates. Marine traffic data for waterways and
harbor systems in the United States are compiled annually by the Corps of Engineers and
published in a document entitled Waterborne Commerce pf jhe. United States The U.S Coast
Guard keeps records of all major accidents/incidents whether or not they result in a loss of
cargo both in regional offices and in a central system in Washington.
Collision and grounding rates have been determined for the seven locations listed in
Table F 2 and may be of use to these locales and others with similar characteristics. Each rate
is derived from actual location-specific data. The rates for any particular body of water or
river, however, may vary from these or national average accident rates as a result of differences
in:
• Traffic density
• Vessel design
• Vessel size
• Vessel speeds
• Local weather conditions
Traffic controls and restrictions, and
« Use of harbor pilots
Where local data exist, they may be used in place of the suggested values given in the
next section.
Suggested Approach for Assessment of Accident Potential
Based upon the information and data presented above, and given the understanding that
harbors and inland waterbodies are of greater concern than shipping activities on the open
ocean and/or otherwise distant from coastlines (in terms of the distances typically associated
with spill effects that may pose a threat to human life and health), accident rates and other spill
characterization factors are presented below for.
• Collisions in lakes, rivers, and intercoastal waterways
• Groundings in lakes, nvers, and intercoastal waterways
• Collisions and groundings in harbors and bays
• Collisions/casualties while vessels are moored/docked
An accident rate of 10-Vmile of travel is suggested for use for collisions in the first
category to cover both the lower expected accident rates on certain slow speed waterways and
the higher ones for congested, highly utilized routes. Based on Gulf Intercoastal Waterway
(ICWW) statistics, a grounding casualty rate of 5 x 10«/ mile is suggested for the serious type
F-ll
-------
TABLE F.2
LOCATION SPECIFIC CASUALTY RATES
Location
Mobile
Houston Ship Channel
Corpus Chnsti
Combined Gulf Ports
Delaware Bay
Providence
Combined Atlantic Ports
Casualty/Mile
1.22 xlO*
7.78 x 10*
1.80 xl(H
9.27 x 1OS
3 61 x 105
1.03 x 10*
3.74 x 105
Collisions and
Groundings/
Harbor Transit
2 89 x 1O3
208xlO3
1.81 x 104
2 11 x 103
1.22 x 1O3
7 06 x 1O*
1.78 x 103
Reference: Abkowitz and Galarraga, 1986
F-12
-------
of grounding which could lead to a release. Note that this is also a "per mile" rate. The
harbor/bay area grounding and collision rate given below is "per transit", while the moored
collision rate is "per port call." (There are two transits per port call.) The suggested rate for
groundings and collisions in a harbor area is lOVtransit, while the suggested casualty rate for
moored or docked vessels is 2 x 104 per port call.
If no distinction is being made with regard to vessel type and construction, it should be
assumed that 0.15 (15%) of accidents result in actual loss of cargo to the environment.
Alternatively, it can be assumed that accidents involving single-hulled vessels result in cargo
loss 0 25 (25%) of the time and that accidents involving double-hulled and bottomed watercraft
result in cargo loss 0 05 (5%) of the time
The recommended distribution of spill amounts is*
• 10% loss of cargo in one tank/compartment ~ 35% of the time
• 30% loss of cargo in one tank/compartment ~ 35% of the time
• Full loss of cargo in one tank/compartment — 30% of the time
This distribution is somewhat more "severe" than those provided by the spill
distributions presented earlier because those distributions are heavily influenced by minor
fitting leaks and the like.
Transportation of Hazardous Materials by Pipeline
Available Accident Data for Pipelines
Failure rates for pipelines are generally given in terms of failures per unit length per year.
Specific data sources may indicate failure rates by product earned, by diameter, and even by
year of installation. This section presents some of the failure rates and other statistics given in
the literature. Values recommended for use in emergency response planning are given in a
later subsection
Two studies have focused on the differences between oil and gas pipelines and onshore
and offshore pipelines (de la Mare and Andersen, 1980 and Andersen and Misund, 1983)
Each concluded that the differences in failure rates showed more of a diameter-dependent
effect (the rate decreases as the diameter increases) than anything else The later study came
up with two basic results ~ an overall pipeline failure rate of about 1 5 x 10-3 per mile per year
for all diameters, and one of about 5 x HH per mile-year for pipelines with a diameter of 20
inches or greater These rates were denved from several studies which all had extremely
consistent results Fourteen years of data for gas transmission and gathering lines in the U S
are also very consistent with these rates (Jones et al, 1986)
F-13
-------
Unlike transportation by marine, truck, or rail which can have accidents/incidents
without a release of hazardous materials, a pipeline failure or accident by definition involves a
release (or partial release) of contents. In the event that there is a failure in a pipeline, most
often the outcome will be a small leak. The leak may persist for a long penod of time,
however, before being detected. This can lead to a larger total volume being released than one
would otherwise have expected.
Historical data on spill size distribution includes an average spill size of 8,060 gal, with
the following spill size breakdown (ADL, 1978):
Gallons
Less than 100
500- 1,000
1,000 - 5,000
10,000 - 25,000
5,000 - 10,000
More than 100,000
23%
15%
8%
23%
8%
23%
The average spill size for polluting incidents was found to be (U S Coast Guard, 1983)-
5,140 gal for oil spills, 350 gal for hazardous substances, and 4,980 gal for all polluting
materials.
The chance of pipelines rupturing has been estimated in a number of studies to be on the
order of 10% - 20%. One study of a propane pipeline cited a figure of 14% (Grolher Baron et
al, 1976). Jones et al (1986) classifies almost 36% of all releases as ruptures, but uses many
definitions of rupture. Very few studies have detailed data on hole size.
It has been estimated that about one-third of natural gas pipeline ruptures ignite, with this
estimate increasing to 45% for very large ruptures. Of the leaks which ignite, a further 30%
are estimated to explode (Kot et al, 1983) Other data sources (including Jones et al, 1986) are
consistent with this or slightly lower.
Use of Local Data
While using national or worldwide pipeline accident rates can accurately estimate the
likelihood of releases from oil or gas pipelines, these rates may overestimate failures for lines
with no internal corrosion possibilities or with many protective design features Likewise,
F-14
-------
failures may be underestimated if there is an extremely corrosive material involved or if other
factors pose a greater than average nsk Using local or state data on lines could reflect some of
these differences, if sufficient information was available
Suggested Approach for Assessment of Accident Potential
Based on the information presented above, an accident rate of 1 5 x lO3/mi-yr is
suggested for lines of unknown size or lines less than 20" in diameter. For pipelines with
diameters greater than or equal to 20", a rate of 5 x lO^/mi-yr is proposed
x
With due consideration to data limitations and the capabilities of pipeline discharge
consequence analysis procedures presented in this guide, the following spill size distribution is
suggested for analyzing pipeline releases of hazardous materials:
• Discharge computed using consequence analysis procedures of Chapter 12
assuming a complete hue break along the route of the pipeline 20%
of the time
• 1 hour release through 1" hole 80% of the time
Handling of Hazardous Materials at Fixed Facilities
Available Accident Data for Fixed Facilities
Most accident data for fixed facilities focuses on incidents in which fatalities occurred,
although polluting incidents are also reported to some level of compliance. More important,
however, is the fact that what information is recorded does not include the necessary exposure
information to determine accident rates Data are presented below to demonstrate the type of
information available. The approach and values recommended for use by emergency planners
are presented in a later subsection.
One very large plant was reported as having a total of 110 spills in 1973 — 23 in the
marine terminal area (with an average spill volume of 2000 gal) and 87 spills in the plant area
(with an average spill volume of 1000 gal). Many of these were contained onsite and corrected
(Carlson et al, 1974).
In the area of loading/unloading spills, very little data is available specifically for rail
cars. Just as for trucks there is a potential for overfills of the receiving storage tank or the rail
car, depending on which is being filled. Trucks, however, are often emptied/filled by the
driver as opposed to plant personnel. The spill distributions given a little farther below
incorporate the results of the reported spills in these areas. While one study found that storage
vessels were about twice as likely to be the cause of injuries and fatalities as other in-plant
events, loading and unloading operations were also cited for causing injuries and fatalities
(Industrial Economics, 1985).
F-15
-------
Upset conditions were cited as being responsible for less than 5% of the release events,
except for process vessels where the percentage rose to 13% (Industrial Economics, 1985)
These figures give credence to the concept of basing accident rates on equipment failure rates
rather than historical release data. COVO (1982) and Considine et al are two sources of some
of the needed failure rate data, many others are also available.
The relative hazards of process versus storage activities have been an ongoing topic of
debate, with the common theory being that process activities are much less safe However, it
has been shown that for large numbers of fatalities (as is the concern of this report) this is not
the case, because of the large volumes generally involved in storage areas (Lees, 1983) This
same study also reported on the frequencies of fires in various types of facilities in 1977.
Refineries > I/year
Natural gas plants 3.7 x 10 Vyear
Tank farms 1.1 x 10 Vyear
Bulk terminals-shore 2.1 x 10 Vyear
Bulk terminals-inland 1.0 x 10 Vyear
Fires are of particular concern in warehouses and other storage situations where they may
be the most likely cause of a hazardous materials release, at least for significant releases
(Forklift handling of various storage containers is very likely to damage packages, but the
damage will generally be quite limited and any releases readily controlled)
An analysis performed by the NFPA Fire Analysis Division showed the following
average annual number of fires, based on 1980 to 1983 data:
Flonst shops, greenhouses 569
Gasoline service stations 1789
Chemical, medical laboratories 172
Agricultural laboratory 22
General research laboratory 733
The exact exposure in each of these categones is not known, but as a rough upper bound it can
be assumed that the average rate of fires per property is 10-3/year (perhaps higher for gasoline
service stations).
Spill size information is widely variable, again reflecting the diversity of facilities falling
into this category. One study (ADL, 1978) found the following distributions.
F-16
-------
Fixed facility
Loading/unloading
-truck
-rail
-marine
Avg.Spill
18,1 10 gal
300
7,390
7,840
<100gal
48%
50%
11%
68%
100-500
20%
50%
33%
20%
500-1000
7%
12%
2%
1000-5000
14%
22%
4%
>5000
11%
22%
6%
The average for marine is very high relative to the distribution because of a few very large
releases
Another study found that the average spill sizes for polluting incidents were (U S Coast
Guard, Calendar Year 1982 and 1983)
Non-Transportation
Spills
Refinery
Bulk storage
Onshore production
Offshore production
Marine Facilities
Fuel transfer
Bulk transfer
Non-bulk transfer
Land Facilities
Oil
400 gal
1,290
4,660
95
310
510
55
400
Hazardous
Substances
12,500
210
—
114
—
1,290
—
100
Total
1,530
1,240
5,660
95
320
500
190
460
Use of Local Data
Local data is obviously critical in terms of determining the exposure, but is not as
important for determining accident rates, because of the approach outlined below. If a
particular facility has completed a nsk assessment for some reason, and makes the results of
this study available to the local emergency response planner, then the study's release estimates
may be replaced for the analysis called out below. The relative rarity of some of the failures
makes it very unlikely that local historical data will include the full range of potential release
events.
F-17
-------
Local or regional data may also be used to determine spill size estimates if insufficient
information is provided by the facilities of concern. EPA regions tend to provide semi-annual
summaries of the releases reported to them
Suggested Approach forAssessment of'AccidentPotential
Based upon the information presented above, the approach suggested for getting a handle
on fixed facility risks is to consider three basic types of release events for plants; one or two
release scenarios for facilities such as water treatment plants, laboratories and industrial
facilities; and one release scenario for warehouses and other facilities storing hazardous
materials. It has been shown that very little specific historical information exists upon which
to base accident rates Hence, the best approach is to look at equipment failure rates. The
increasing use of physical barriers to limit spills, drainage systems to channel spills and venting
and scrubbing systems to control releases all help to render this simplified accident estimation
procedure more meaningful.
For example, a large facility may be coarsely modelled as having storage operations,
loading/unloading operations, and processing operations. These can respectively be
represented by storage tank failures and leaks, hose failures, and piping and process vessel
failures. The rates suggested for each of these are:
Storage tank - double walled lOVtank-year
Storage tank - single walled lOYtank-year
Pressure vessels ICH/vessel-year
Piping 1.5 x 10-Vft-year
Loading hoses KH/operation or
lOVhose-year
While these certainly do not cover all potential release scenarios, they do capture some of
the more likely ways to lose large volumes of material. The only piping of prune concern is
that of relatively large diameter and long segments In other words, a 100-foot expanse of 8"
pipe should be counted if it contains hazardous materials, but not 10 or 20 foot sections
between vessels. The spill size is generally taken to be a function of the specific release
scenario since historical averages are difficult to apply.
For the middle category of industrial users, water treatment plants, laboratories, etc., the
main focus should be on storage tank or container failures. Piping failures or loading hose
failures may be considered if there is a significant amount of piping (say over 100 feet) or if
there are a lot of loadings/unloadings (say 10 or more per year). The rates to be used are the
same as those listed above.
F-18
-------
Storage of hazardous materials, such as in warehouses or greenhouses, may also result in
failures of storage containers, but the greater threat here is probably from a fire which spreads
to the storage area and results in release, ignition, explosion, and/or combustion of stored
materials (with attendant evolution of potentially toxic smoke) The occurrence rate of such
fires is suggested to be lOVyr, with a release of 10% to 100% of the stored volume This is
one area in which more specific local data and information would be particularly helpful for
better definition of scenarios and estimation of then- likelihood
The approach of this section is not to cover all potential sources of hazardous materials
releases, but to cover some of the larger and more likely events
References
Abkowitz, M. and J. Galarraga. "Tanker Accident Rates and Expected Consequences in
U.S. Ports and High Seas Regions," Recent Advances in Hazardous Materials
Transportation Research, An International Exchange, State-of-the-Art Report 3,
Transportation Research Board, Washington, DC, 1986, pp 161-169.
American Petroleum Institute. "Summary of Motor Vehicle Accidents in the Petroleum
Industry for 1982," June 1983.
Andersen, T. and A. Misund. "Pipeline Reliability. An Investigation of Pipeline Failure
Characteristics and Analysis of Pipeline Failure Rates for Submarine and
Cross-Country Pipelines," Journal of Petroleum Technology, Vol. 35, No. 4, April
1983, pp 709-717
Arthur D Little, Inc "Risk Analysis of Ships Serving Expanded Mongstad Refinery,"
Report to Statoil, May 1983
Arthur D. Little, Inc. "Assessment of Risks and Risk Control Options Associated with
Liquefied Natural Gas Trucking Operations from Distngas Terminal, Everett,
Massachusetts," prepared for the U S Department of Transportation, Contract No
DOT-RC-82037, June 1979.
Arthur D Little, Inc. "Estimation of the Frequency and Costs Associated with the Cleanup
of Hazardous Materials Spills," prepared for the US Environmental Protection
Agency, Contract No 68-01-3857, September 1978
Association of American Railroads "Railroad Facts, 1985 Edition," August 1985.
Carlson, L E, J F. Erdmann and G J Hanks, Jr "How One Chemical Company is Attacking
the Spill Problem," Proceedings of 1974 National Conference on Control of
Hazardous Materials Spills, pp 106-118.
F-19
-------
Chemical Week. "Getting railroad safety back on track," March 15,1978, pp 20-21.
Clarke, R.K. et al "Seventies of Transportation Accidents," Sandia Laboratories, NTIS
SLA-74-0001, July 1976.
Considine, M., G.C. Grint and P.L. Holden "Bulk Storage of LPG - Factors Affecting
Offsite Risk," Institution of Chemical Engineers Symposium Series No. 71, pp.
291-320.
COVO Steering Committee. Risk Analysis of Six Potentially Hazardous Industrial Objects
in the Rijnmond Area, a Pilot Study, Boston' D Reidel Publishing Co, 1982.
Croce, P.A. et al "A Feasibility Study of a Sealed Safety Monitor for Trucks Carrying LNG
and Other Hazardous Materials," prepared by Arthur D. Little, Inc for the U S
Department of Energy, Office of Operational Safety Programs, DOE/EV/10502-1,
December 1982.
de la Mare, R.F. and O. Andersen Pipeline Reliability, Report No 80-0572, Det Norske
Veritas, August 1980.
Dennis, A.W. et al. "Seventies of Transportation Accidents Involving Large Packages,"
Sandia Laboratories, NTIS SAND-77-0001, May 1978.
Elder, H.K. et al. "An Assessment of the Risk of Transporting Spent Nuclear Fuel by
Truck," prepared by Pacific Northwest Laboratory for the U.S. Department of Energy,
PNL-2588, November 1978
Federal Railroad Administration. "Accident/Incident Bulletin, No. 156, Calendar Year
1987," July 1988.
Federal Railroad Administration. "Accident/Incident Bulletin, No 153, Calendar Year
1984," June 1985.
Federal Railroad Administration. "Accident/Incident Bulletin, No 152, Calendar Year
1983," June 1984.
Federal Railroad Administration "Accident/Incident Bulletin, No 151, Calendar Year
1982," June 1983.
Geffen, C A. et al. "Assessment of the Risk of Transporting Propane by Truck and Train,"
prepared by Pacific Northwest Laboratories for the US Department of Energy,
PNL-3308, March 1980
F-20
-------
Grollier Baron, R, R. Lombre and V. Rochina. "Hazard Analysis for a Liquid Propane
Pipeline," Recent Advances in Hazardous Materials Transportation Research, An
International Exchange, State-of-the-Art Report 3, Transportation Research Board,
Washington, DC, 1986, pp. 141-146.
Harvey, A.E., P.C. Conlon, and T S. Glickman, "Statistical Trends in Railroad Hazardous
Materials Transportation Safety - 1978 to 1986," Publication R-640, Association of
American Railroads, Washington Systems Center, September 1987.
ICF, Inc. "Assessing the Releases and Costs Associated with Truck Transport of Hazardous
Wastes," prepared for the U.S. Environmental Protection Agency, NTIS PB84-224468
1984.
Ichmowski, T. "New measures to bolster safety in transportation," Chemical Engineering,
November 12,1984, pp. 35-39.
Industrial Economics, Inc. "Acute Hazardous Events Data Base," prepared for the
Environmental Protection Agency, NTIS PB86-158946, December 1985.
Insurance Institute for Highway Safety, "Big Tracks and Highway Safety," 1985.
Jones, E J et al. "An Analysis of Reportable Incidents for Natural Gas Transmission and
Gathering Lines 1970 through 1984," prepared by Battelle Columbus Laboratories for
the American Gas Association, NG-18 Report No. 158, March 1986
Kloeber, G. et al. "Risk Assessment of Air Versus Other Transportation Modes for
Explosives and Flammable Cryogenic Liquids, Volume I: Risk Assessment Method
and Results," prepared by ORI, Inc. for Materials Transportation Bureau, NTIS
PB80-138472, December 1979.
Kot, C.A. et al. "Hazardous Material Accidents Near Nuclear Power Plants: An Evaluation
of Analyses and Approaches," prepared by Argonne National Laboratory for the U.S.
Nuclear Regulatory Commission, NUREG/CR-3548, October 1983
Lees, F.P. "The Relative Risk From Materials in Storage and in Process," Jorunal of
Hazardous Materials, 8,1983, pp. 185-190.
Maritime Transportation Research Board "Reducing Tankbarge Pollution," Report of the
Committee on Reducing Tankbarge Pollution, National Research Council, June 1981.
Midwest Research Institute, "Present Practices of Highway Transportation of Hazardous
Materials, Task B Interim Report, Literature Review," prepared for Federal Highway
Administration, DTFH61-86-C-00039, January 30,1987.
F-21
-------
Midwest Research Institute, "Present Practices of Highway Transportation of Hazardous
Materials, Task C Interim Report, Analysis of Existing Data Bases," prepared for
Federal Highway Administration, DTFH61-86-C-00039, May 19,1987
National Safety Council. "Accident Facts, 1988 Edition "
Nayak, P.R. "An Analysis of the Probability of Derailment with a Train Consist," prepared
by Arthur D. Little, Inc. for the Transportation Systems Center, U.S. Department of
Transportation, December 1979
Nayak, P R., D.B. Rosenfield and J H Hagopian. "Event Probabilities and Impact Zones for
Hazardous Materials Accidents on Railroads," prepared by Arthur D. Little, Inc for
the Federal Railroad Administration, DOT/FRA/ORD-83/20, November 1983.
Office of Technology Assessment. "Transportation of Hazardous Materials," OTA-SET-340,
Washington, DC: U.S. Government Printing Office, July 1986.
Rhoads, RE. et al. "An Assessment of the Risk of Transporting Gasoline by Truck,"
prepared by Pacific Northwest Laboratory for the U.S Department of Energy,
PNL-2133, November 1978.
Robison, R.W. "Hazardous Materials Hazard Analysis," prepared by the City of Portland,
Office of Emergency Management, June 1981
Skolnick, M. Discussion comments on paper by D.J. Friend, "Public Concerns and
Hazardous Matgenals Transportation Safety: Closing the Gap," Recent Advances m
Hazardous Materials Transportation Research, An International Exchange,
State-of-the-Art Report 3, Transportation Research Board, Washington, DC, 1986, pp
193-194.
Smith, R.N. and EJL. Wilmot. "Truck Accident and Fatality Rates Calculated from
California Highway Accident Statistics for 1980 and 1981," prepared by Sandia
National Laboratories for U S. Department of Energy, SAND-82-7066, November
1982
Stewart, R.J. and M B. Kennedy. "An Analysis of U S. Tanker and Offshore Petroleum
Production Oil Spillage Through 1975," prepared by Martingale, Inc for the U S.
Department of the Interior, May 1978.
Todd Shipyards. "Nature of Ship Collisions Within Ports," Report for U.S. Maritime
Administration, NTIS PB-255 304, April 1976.
F-22
-------
U.S. Coast Guard. "Polluting Incidents In and Around U.S. Waters, Calendar Year 1982 and
1983."
U.S. Coast Guard. Draft Regulatory Analysis and Environmental Impact Statement for
Design Standards for New Tank Barges and Regulatory Action for Existing Tank
Barges to Reduce Oil Pollution Due to Accidental Hull Damage. Docket #75-083 and
#75-083a, Office of Merchant Marine Safety, May 1979.
Urbanek, G.L. and E.J. Barber. "Development of Criteria to Designate Routes for
Transporting Hazardous Materials," prepared by Peat, Marwick, Mitchell and Co. for
the Federal Highway Administration, NTIS PB81-164725, September 1980.
von Herberg, P. "Prevention is the Best Cure," Chemical Purchasing, September 1979, pp.
79-84.
F-23
•6 US GOVERNMENT PRINTING OFFICE 1980 725-600/20672
-------
------- |